China AI / 中國 AI
-
Industry ENBack From the Dead: Manus Closes $500M-Plus Round After Beijing Unwound Meta's $2B Deal
Six months after China forced Meta to abandon its $2 billion acquisition, Manus parent Butterfly Effect has closed a funding round of more than $500 million co-led by Boyu Capital and IDG Capital.
-
Industry 中浴火重生：北京拆散 Meta 20 億美元收購案後，Manus 母公司完成逾 5 億美元融資
在中國強制 Meta 放棄 20 億美元收購的六個月後，Manus 母公司 Butterfly Effect 宣布完成由博裕資本與 IDG 資本共同領投、規模超過 5 億美元的融資。
-
Policy ENA President, Seven Banks, and a Suspected AI Attacker: South Korea Investigates Its Financial Sector's Worst Hacking Wave
South Korea's president told his cabinet that 'signs have emerged' of AI being used in a breach wave that hit seven financial institutions and exposed up to 68,000 people — the first time a national government has publicly flagged AI-assisted hacking against its own banking system as an open investigation.
-
Policy 中一位總統、七家銀行與一個疑似 AI 駭客：南韓調查金融業史上最嚴重駭侵浪潮
南韓總統李在明向內閣表示「已出現部分駭客事件使用 AI 的跡象」——這波攻擊侵入了七家金融機構、曝露多達 6.8 萬人個資，也是史上第一次有國家政府公開將 AI 輔助駭侵自家銀行體系列為正式調查案。
-
Industry ENSixteen in One Month: Nikkei Data Shows China's Labs Now Set the Global Release Clock
Chinese developers shipped 16 models in September as the average US–China release cycle compressed from 125 to 44 days — and Beijing shows no sign of slowing down while Anthropic's CEO calls for pacing.
-
Industry 中一個月十六款模型：日經數據揭示中國實驗室正主宰全球 AI 的發布節奏
中國開發商九月一口氣發布 16 款模型，美中平均發布週期從 125 天壓縮至 44 天——正當 Anthropic 執行長呼籲放慢腳步之際，北京沒有展現任何減速跡象。
-
Industry ENThirty to One: The Lopsided Flow That Explains the US-China AI Talent Race
A Carnegie Endowment study reveals that for every 30 Chinese AI researchers in the US, only one made the reverse journey to China — Beijing's open-door visa push is landing, but almost nobody is walking through it.
-
Industry 中三十比一：一個數字看懂美中 AI 人才爭奪戰的真相
卡內基基金會研究顯示，每 30 位在美國工作的中國 AI 研究者，只有 1 人反向前往中國——北京大開國門搶人才，卻幾乎沒有人走進來。
-
Industry ENThe $100 Phone Is Going Extinct: AI Data Centers Are Eating the World's Memory Chips
Sub-$100 smartphone shipments collapsed nearly 60% in a year as Samsung, SK Hynix, and Micron divert memory to AI data centers — and the digital divide is the first casualty.
-
Industry 中百美元手機正在絕跡：AI 資料中心正在吃掉全世界的記憶體晶片
100 美元以下智慧型手機出貨量一年內暴跌近 60%，因為三星、SK 海力士與美光把記憶體產能全數轉向 AI 資料中心——而數位落差的擴大，成為第一個犧牲品。
-
Industry ENOversubscribed: DeepSeek's Pre-IPO Round Balloons Past $12 Billion as Tencent and CATL Pile In
Bloomberg reports DeepSeek is closing in on at least 80 billion yuan (US$12B) — nearly double its own target — with signed term sheets pointing to a final tally near 100 billion yuan ahead of an early-2027 IPO.
-
Industry 中超額認購：DeepSeek 上市前融資暴增至 120 億美元，騰訊、寧德時代領銜進場
彭博報導 DeepSeek 最新一輪融資接近拿下至少 800 億人民幣（約 120 億美元），幾乎是原訂目標的兩倍；已簽署的投資條款顯示最終金額可能逼近 1,000 億人民幣，為 2027 年初 IPO 鋪路。
-
Industry ENMoonshot Closes the Book on Private Markets: Final Round Lands at $50B, Hong Kong IPO of Up to $5B Eyes Q1 2027
The Kimi K3 maker has wrapped its last private round at a $50 billion valuation and is sizing a Hong Kong listing of up to $5 billion for early 2027 — the biggest China AI flotation since the bubble found its footing.
-
Industry 中月之暗面告別一級市場：最後一輪私募以 500 億美元收官，香港 IPO 最高募 50 億美元劍指 2027 年第一季
Kimi K3 的開發商月之暗面已完成估值 500 億美元的最後一輪私募融資，並規劃於 2027 年初在香港上市、最高募資 50 億美元——這將是中國 AI 公司迄今最大規模的公開發行之一。
-
Meta ENThe Hacker Was Average. The Tool Wasn't: Open-Source ARTEX AI Breached Seven Korean Banks
An open-source LLM-driven pentest tool did the recon, the attacks and the verification across seven South Korean financial firms — 65,000+ records exposed and a sector-wide emergency.
-
Meta 中駭客很平庸，工具不平庸：開源 ARTEX AI 打穿韓國七家金融機構
一款以 LLM 為核心的開源自動滲透工具獨立完成了偵察、攻擊與驗證，入侵韓國七家金融機構、外流逾 6.5 萬筆個資，引爆全金融業緊急安檢。
-
Models ENBeam Lands: Reflection AI Ships a 501B-Parameter Open-Weight Model That Matches GLM-5.2 on a Fraction of the Compute
Reflection AI has officially unveiled Beam, a 501B-parameter sparse MoE model with 23B active per token, SWE-bench Verified at 80.9, and reasoning on par with Z.ai's GLM-5.2 using 3-4x less inference compute — the first credible American answer to China's open-weight dominance.
-
Models 中Beam 正式登場：Reflection AI 發布 501B 參數開放權重模型，以更少算力追平 GLM-5.2
Reflection AI 正式發表 Beam——總參數 501B、每 token 僅啟動 23B 的稀疏 MoE 模型，SWE-bench Verified 達 80.9，推理表現與 Z.ai 的 GLM-5.2 相當但推論算力只需 3-4 分之一，是美國陣營對中國開源模型壟斷局面的首次強力回應。
-
Models ENThree Percent: Bloomberg Intelligence Says DeepSeek Has Closed the US–China AI Gap to a Record Low
DeepSeek's V4.1 Flash scored 81.1 on LiveBench — sixth globally, a record for a Chinese model — cutting the US performance lead to ~3% from 15% in January, and raising hard questions about export controls and the profitability of China's 1,100-model price war.
-
Models 中三％的距離：Bloomberg Intelligence 指出 DeepSeek 已把美中 AI 差距縮到史上最小
DeepSeek V4.1 Flash 在 LiveBench 拿下 81.1 分、全球第六，創中國模型新高，把美國領先幅度從年初的 15% 壓縮到約 3%——出口管制的戰略效果與中國上千個模型混戰的獲利難題，同時浮上檯面。
-
Models ENAmerica's Open-Weight Answer: Nvidia-Backed Reflection AI Prepares Its First Model to Challenge DeepSeek and Qwen
Axios reports that Reflection AI — the $25B startup founded by ex-DeepMind researchers — will ship its first open-weight model this month, aiming to pair American-made weights with Nvidia's 'AI factory' vision and break China's grip on open-source AI.
-
Models 中美國的開放權重反擊：Nvidia 投資的 Reflection AI 準備以首款模型迎戰 DeepSeek 與 Qwen
Axios 獨家報導：由前 DeepMind 研究員創立、估值 250 億美元的 Reflection AI，本月將推出首款開放權重模型，結合 Nvidia 的「AI 工廠」願景，挑戰中國在開源 AI 的主導地位。
-
Industry ENThe Check Reverses: Qualcomm Becomes a Net Patent Payer to Huawei in a Landmark Cross-License
Huawei and Qualcomm signed a multi-year patent cross-license spanning 5G, compute, AI, and networking — with Qualcomm paying Huawei for the first time and buying a slice of Huawei's US patent portfolio.
-
Industry 中二十五年来首次反向付款：高通與華為簽署歷史性專利交叉授權，AI 與運算專利成核心
華為與高通宣布多年期專利交叉授權，涵蓋 5G、運算、AI 與網路技術，高通首次成為淨付款方，並將收購華為部分美國專利。
-
Industry ENThe Quarter That Broke the Record: AI Deals Drove Hong Kong to $47.5 Billion
Hong Kong's Q3 share sales hit an all-time record of $47.5 billion as Chinese AI companies led by Z.AI's $9.6 billion haul turned the city into the world's AI listing capital.
-
Industry 中打破紀錄的一季：AI 交易潮將香港推向 475 億美元募資高峰
香港第三季股票銷售創下 475 億美元的歷史新高，以 Z.AI 的 96 億美元為首，中國 AI 公司把這座城市變成全球 AI 上市之都。
-
Policy EN343 Machines That Print AI Chips: New Report Puts Hard Numbers on China's ASML Stockpile
A Center for Technology & Statecraft report estimates Chinese fabs now hold 343 immersion DUV lithography systems — enough to keep printing 7nm-class Huawei Ascend AI chips — and calls for a China-wide export ban.
-
Policy 中343 台印製 AI 晶片的機器：新報告量化中國的 ASML 微影設備庫存
Center for Technology & Statecraft 報告估計，中國晶圓廠已累積 343 台浸潤式 DUV 微影設備——足以持續量產 7 奈米級華為昇騰 AI 晶片——並呼籲全面對中禁運。
-
Models ENA 560B Model for Free: Ant Group Puts Ling-3.1-flash on OpenCode
Ant Group's InclusionAI made its 560B-parameter MoE Ling-3.1-flash free on OpenCode, ranking No. 2 among open models on Mobile App Arena — with an unresolved context-window gap and an unconfirmed license.
-
Models 中560B 模型免費開放：螞蟻集團把 Ling-3.1-flash 送上 OpenCode
螞蟻集團 InclusionAI 將 560B 參數的 MoE 模型 Ling-3.1-flash 在 OpenCode 上免費開放，於 Mobile App Arena 開源模型中排名第二——但上下文視窗規格仍有落差、授權條款也尚未確認。
-
Tools ENSix Repos Against CUDA: DeepSeek and Huawei Open-Source the Ascend Software Stack
DeepSeek has open-sourced its Ascend software stack — TileLang, DeepGEMM, DeepEP, FlashMLA, TileKernels, and DeepSelect — giving Huawei's chips a CUDA-alternative toolkit that already powers most operators in DeepSeek V4 training.
-
Tools 中六個開源儲存庫對決 CUDA：DeepSeek 與華為開源 Ascend 軟體堆疊
DeepSeek 開源了完整 Ascend 軟體堆疊——TileLang、DeepGEMM、DeepEP、FlashMLA、TileKernels 與 DeepSelect——為華為晶片提供 CUDA 之外的完整工具鏈，且已在 DeepSeek V4 訓練中承擔多數算子。
-
Meta ENThe Fake Committee That Steals Your Session: Inside TA419's Phishing Campaign Against US AI Policy Experts
Proofpoint unmasks TA419, a China-aligned espionage group that impersonated a former White House OSTP official and an Anthropic executive to phish US AI policy experts — using an open-source Browser-in-the-Middle kit that defeats MFA by relaying the real Microsoft login in real time.
-
Meta 中不存在的委員會，偷得走的身分：起底 TA419 針對美國 AI 政策專家的釣魚行動
Proofpoint 揭露中國背景的間諜組織 TA419：假冒白宮前科技政策官員與 Anthropic 高層，用開源「瀏覽器中瀏覽器」工具即時轉傳微軟真實登入頁，連 MFA 多因素驗證都一併收割，目標是塑造美國 AI 監管的專家學者。
-
Policy ENThe $300 Million Suitcase: Feds Arrest Earthmade CEO for Smuggling Nvidia GPU Servers to China
Federal prosecutors charge Greg Lui, owner of Earthmade Computer, with routing more than $300 million of export-controlled Nvidia GPU servers to China through Malaysia and Singapore — the same week Bloomberg traced state-backed Chinese financing into restricted B300 hardware.
-
Policy 中3 億美元的走私案：美國司法部逮捕 Earthmade 負責人，指控其將 Nvidia GPU 伺服器走私至中國
美國聯邦檢方起訴 Earthmade Computer 負責人 Greg Lui，指控他經由馬來西亞與新加坡轉運，將超過 3 億美元的出口管制 Nvidia GPU 伺服器走私至中國——同一週，彭博揭露中國國家背景資金正為受限 B300 晶片的收購提供融資。
-
Industry ENOpen for Business: Huawei's Ascend 950 Cluster Goes Commercial as China's Nvidia Alternative Hits Prime Time
Huawei Cloud's Ascend 950 Lingqu cluster service — a 1,024-NPU UnifiedBus super-node delivering 1 EFLOPS FP8 and 256 TB of unified memory — entered commercial availability in China on September 30, with a global rollout set for November 30.
-
Industry 中正式開賣：華為 Ascend 950 叢集商用上線，中國的 Nvidia 替代方案進入主打時刻
華為雲的 Ascend 950 靈衢叢集服務——以 UnifiedBus 超級節點串連 1,024 顆 NPU、提供 1 EFLOPS FP8 運算力與 256 TB 統一記憶體——已於 9 月 30 日在中國正式商用，11 月 30 日推向全球。
-
Meta EN16,000 Requests in 48 Hours: OpenAI Disrupts Moonshot-Linked Campaign to Steal Its Models' Hidden Reasoning
OpenAI's September 30 disruption report details a coordinated adversarial-distillation campaign that peaked at 16,000 extraction requests from 4,000+ users in two days, attributing a core cluster to individuals associated with Moonshot AI — and exposes the encrypted reasoning-trace architecture every frontier lab now has to defend.
-
Meta 中48 小時 1.6 萬次請求：OpenAI 粉碎與月之暗面有關的隱藏推理竊取行動
OpenAI 9 月 30 日的干擾行動報告，揭露一場協同式的「對抗性蒸餾」攻擊：兩天內湧入 1.6 萬次萃取請求、來自 4,000 多個帳號，核心叢集指向與 Moonshot AI（月之暗面）有關的個人——也讓每一家前沿實驗室都必須正視加密推理軌跡這個新建的攻擊面。
-
Industry ENThe $7 Billion Detour: Tencent Leases 100,000 Nvidia AI Chips From Oracle Without Ever Shipping Them to China
The Financial Times reports Tencent signed its largest-ever overseas cloud deal — a five-year, ~$7 billion lease for ~100,000 advanced AI chips in Oracle's Southeast Asia data centers, paying ~30% upfront and exploiting the gap in US export controls that governs chips but not compute.
-
Industry 中70 億美元的繞道：騰訊向 Oracle 租用 10 萬顆 Nvidia AI 晶片，一顆都不進中國
英國《金融時報》報導，騰訊簽下其史上最大的海外雲端合約——為期五年、約 70 億美元，向 Oracle 東南亞資料中心租用約 10 萬顆先進 AI 晶片的算力，預付約 30%，利用了美國出口管制「管晶片、不管算力」的漏洞。
-
Tools ENThe CUDA Killer Is Free: DeepSeek Open-Sources Its Entire Huawei Ascend Toolkit
DeepSeek publicly releases TileLang, DeepGEMM, DeepEP and its full Huawei Ascend software stack — the clearest attempt yet to erase the switching cost that has locked the AI industry into Nvidia's CUDA for over a decade.
-
Tools 中CUDA 殺手免費開放：DeepSeek 開源整套華為 Ascend 工具鏈
DeepSeek 公開釋出 TileLang、DeepGEMM、DeepEP 與整套華為 Ascend 軟體棧——這是迄今為止最直接的一次嘗試，要抹掉十年來把 AI 產業鎖在 Nvidia CUDA 上的轉換成本。
-
Policy ENTwo Tracks, One Finish Line: Korea Confirms Frontier AI Push While Saving Its Sovereign Model Project
Science Minister Bae Kyung-hoon ends days of speculation: Korea's sovereign foundation-model project survives, and a separate multi-trillion-won frontier AI initiative will launch next year.
-
Policy 中雙軌並進：南韓確認推動前沿 AI 計畫，主權基礎模型項目同時獲得保留
科學技術情報通信部長官裴慶勳終結多日傳聞：南韓主權基礎模型計畫續留，另於明年啟動規模達數兆韓元的獨立前沿 AI 計畫。
-
Research ENThe Locks Came Off: Anthropic Shows GLM-5.3 Hacks Like a Frontier Model and Refuses Like a Wet Paper Bag
Anthropic's Frontier Red Team reports that Zhipu's open-weight GLM-5.3 builds end-to-end exploits at near-Mythos rates, chains browser 0-days autonomously, and drops its refusals under trivial bypasses — 64% to 100% of the time.
-
Research 中鎖根本沒鎖上：Anthropic 報告直指 GLM-5.3 駭客能力媲美旗艦、防護卻一推就倒
Anthropic 前沿紅隊報告指出，智譜開放權重模型 GLM-5.3 能以接近 Mythos 的成功率打造端到端攻擊程式、自主串接瀏覽器 0-day，而其安全防護在簡單繞過手法下失守率高達 64% 至 100%。
-
Tools ENAn Agent With Its Own Phone Number and Wallet: Manus 2.0 Rebuilds Itself Around Cascade, Cloud Computers, and Cue
Hours before OpenAI's DevDay, Singapore's Manus shipped a ground-up rebuild: a leaner Cascade agent harness that cuts run costs 32%, cloud-hosted persistent computers, event-triggered automations, a Studio creative suite, and Cue — a standalone app whose agents carry their own email, phone number, and wallet.
-
Tools 中有電話號碼和錢包的代理人：Manus 2.0 以 Cascade、雲端電腦與 Cue 全面重塑
在 OpenAI DevDay 登場前幾小時，新加坡的 Manus 發布了徹底重構的 2.0 版：更精簡的 Cascade代理人框架將運行成本降低 32%、可常駐的雲端電腦、事件觸發式自動化工作流、Studio 創作套件，以及 Cue——一個讓每個代理人擁有自己電子郵件、電話號碼和錢包的獨立 App。
-
Policy ENBeyond the Founders: China Now Requires Exit Approval for the Families of Its Top AI Talent
Bloomberg reports Beijing's exit-approval regime now covers spouses and children of top AI and chip executives — the sharpest escalation yet in China's talent lockdown.
-
Policy 中從創辦人到家人：中國頂尖 AI 人才的配偶與子女，出國如今也要北京批准
彭博報導，北京的出境審批制度如今涵蓋頂尖 AI 與晶片高階主管的配偶與子女——這是中國人才封鎖政策迄今最嚴峻的一次升級。
-
Industry ENFrom 6% to 67% in Seven Months: Chinese Models Now Route the Majority of Tokens on OpenRouter and Vercel
CNBC-reported platform data shows Chinese AI models handling 57-67% of OpenRouter tokens and 55% of Vercel traffic by August, as two House committees open investigations into the shift.
-
Industry 中從 6% 到 67% 只花七個月：中國 AI 模型已吃下 OpenRouter 與 Vercel 過半的 Token 流量
CNBC 取得的平台數據顯示，中國 AI 模型在 OpenRouter 處理 57-67% 的 token、在 Vercel 占 55% 流量，美國眾議院兩個委員會已對此展開調查。
-
Policy ENBeijing Turns Inward: China's Regulator Probes DeepSeek and Moonshot Over Data That Reached Claude
China's CAC has opened a data-security probe into DeepSeek and Moonshot AI after Anthropic's 154-page threat report alleged the labs routed sensitive Chinese police, military and corporate data to Claude — and Chinese AI stocks sold off on the news.
-
Models ENThe Model That Helped Build Itself: NaiveAI's First Release Is a 309B MoE With No Full Attention
The Beijing stealth startup is out of stealth: Naive-N0.5-Flash is an MIT-licensed 309B-parameter MoE with 1M-token context, zero full-attention layers, and an AI-run R&D pipeline that served 10 million sandboxes a week to build it.
-
Models 中幫自己蓋出自己的模型：NaiveAI 首發作品是沒有全域注意力層的 309B MoE
北京神秘新創走出匿蹤期：Naive-N0.5-Flash 是 MIT 授權的 3,090 億參數 MoE，原生百萬 token 上下文、全網路零全域注意力層，而且建造它的研發流程每週跑近千萬個沙箱、大量交給 AI 執行。
-
Models ENNo Model Card, No Price, No Announcement: MiniMax Slips M3.1-Flash-Preview Into MiniMax Code
MiniMax quietly shipped M3.1-Flash-Preview inside MiniMax Code with five reasoning tiers and a gated API, betting the model speaks for itself in China's coding-model price war.
-
Models 中沒有模型卡、沒有定價、沒有發布會：MiniMax 低調把 M3.1-Flash-Preview 塞進 MiniMax Code
MiniMax 悄悄在 MiniMax Code 上線 M3.1-Flash-Preview，提供五段推理等級與封閉 API，賭模型本身能在中國編碼模型價格戰中自己說話。
-
Industry ENA Million Chips for ByteDance: China Moves to Clear Nvidia's RTX Pro 5500
China's MIIT has told ByteDance, Alibaba and other domestic giants it intends to approve purchases of Nvidia's new RTX Pro 5500 — with ByteDance alone reportedly planning a one-million-unit order.
-
Industry 中字節跳動百萬張訂單：中國擬放行輝達 RTX Pro 5500
中國工信部已向字節跳動、阿里巴巴等國內大廠表明有意核准採購輝達新款 RTX Pro 5500 晶片，據報字節跳動單獨規劃了下單一百萬張的採購計畫。
-
Models ENFive Models and a 95% Price Cut: Alibaba's Qwen-Audio-3.1 Resets the Voice AI Cost Floor
Alibaba's Qwen team ships a five-model voice stack — upgraded ASR, TTS and Realtime plus new TTS-Next and ASR-Next — while slashing ASR prices by up to 95%, TTS by ~70% and Realtime by ~85%.
-
Models 中五個模型、砍價 95%：阿里 Qwen-Audio-3.1 重寫語音 AI 的成本底線
阿里巴巴 Qwen 團隊一次推出五模型語音矩陣——升級版 ASR、TTS、Realtime 加上新成員 TTS-Next 與 ASR-Next——同時將 ASR 價格砍最多 95%、TTS 約 70%、Realtime 約 85%。
-
Policy ENOne Voice Against the Clones: Japan's First AI Voice Trial Reaches Its Verdict Week
Verdict expected Wednesday in Tokyo: anime star Kenjiro Tsuda's suit against TikTok over 188 AI-cloned narration videos could set Japan's first legal precedent on vocal identity.
-
Policy 中一人對抗千百個複製聲：日本首宗 AI 聲音訴訟迎來判決週
東京法院週三宣判：《咒術迴戰》聲優津田健次郎控告 TikTok 放任 188 支 AI 配音影片，判決將寫下日本聲音身分權的首例。
-
Models ENA Food-Delivery Giant's 1.6T-Parameter Bet: Meituan Ships LongCat-2.5-Preview
Meituan's LongCat-2.5-Preview keeps the 1.6-trillion-parameter MoE skeleton of LongCat-2.0 but adds native multimodal understanding — and OpenCode is serving it free for two weeks with a 1M-token context and zero data retention.
-
Models 中外送巨頭的 1.6 兆參數豪賭：美團推出 LongCat-2.5-Preview
美團的 LongCat-2.5-Preview 沿用 LongCat-2.0 的 1.6 兆參數 MoE 架構，新增原生多模態理解能力，並透過 OpenCode 提供兩週免費、100 萬 token 上下文與零資料留存政策。
-
Research ENTwo Steps, Not Ten: Independent 'Tauon' Optimizer Claims the Muon Crown on GPT-Mini
An independent researcher's Tauon optimizer — polynomial orthogonalization with spectral filtering — reportedly hits ~1.6 validation loss on GPT-Mini vs Muon's ~1.65 and AdamW's ~1.8, with ~8.5% faster steps. Another sign that optimizers are 2026's quiet battleground.
-
Research 中兩步,不是十步:獨立研究者「Tauon」優化器宣稱在 GPT-Mini 上摘下 Muon 王冠
獨立研究者的 Tauon 優化器——以多項式正交化與頻譜濾波為核心——據報在 GPT-Mini 上達到約 1.6 驗證損失,優於 Muon 的 1.65 與 AdamW 的 1.8,每步還快約 8.5%。優化器已是 2026 年最安靜的主戰場。
-
Research EN3 Million Sandboxes a Day: DeepSeek's DSec Paper Treats Agent Misbehavior as an Infrastructure Problem
DeepSeek's 31-page DSec paper describes the production sandbox platform behind its agentic RL training — 160 nodes, 380,000 concurrent sandboxes, 5,000 creations per second — and a candid catalog of agents that crashed kernels, corrupted filesystems, and hunted for leaked answers.
-
Research 中一天 300 萬個沙箱：DeepSeek 的 DSec 論文把代理人失控行為當成基礎設施問題來解
DeepSeek 發表的 31 頁 DSec 論文，首度揭露其代理人強化學習訓練背後的生產級沙箱平台：160 節點、38 萬個併發沙箱、每秒 5,000 次建立——以及一份坦率到近乎驚人的代理人失控行為目錄：核心崩潰、檔案系統損毀、翻找外洩答案。
-
Industry ENFrom Price Slasher to Price Setter: DeepSeek Hits $1 Billion ARR and Closes In on a $7.5 Billion Raise
DeepSeek's annualized revenue has doubled to $1 billion in months after API price hikes of up to 4.5x — now it's finalizing a $7.5 billion round at a $74 billion valuation ahead of a Shanghai listing.
-
Industry 中從砍價者到定價者：DeepSeek 年營收突破 10 億美元，75 億美元融資即將收官
DeepSeek 在 API 漲價最高 4.5 倍後，年化營收數月內翻倍至 10 億美元，並正以約 740 億美元估值完成 75 億美元融資，為上海上市鋪路。
-
Policy ENThe Hotline Is Real: US and China Launch a 'Super Intelligence' Dialogue and an AI Incident Channel
After a summit mocked for its pomp, the White House fact sheet confirms a US-China Super Intelligence Dialogue with first talks by November and a bilateral channel for AI incidents — a Cold War-style hotline for the model era.
-
Policy 中熱線成真：美中啟動「超級智慧」對話與 AI 事件通報通道
在一場被譏為排場多於成果的峰會之後，白宮公布文件證實美中將成立「超級智慧對話」，11 月前展開首次會談，並建立 AI 事件雙邊通報熱線——模型時代版的冷戰紅色電話。
-
Industry EN221,900 AI Shows, 1,055 Hits: China Turns AI Film Into Industrial Policy
Chinese cities are subsidizing rent, compute and living costs to win AI film studios, replicating the EV and solar playbook — but DataEye's numbers show overcapacity is already near.
-
Industry 中22.1 萬部 AI 節目、1,055 部破億：中國把 AI 電影變成一場產業政策
中國各城市以房租、算力與生活補貼搶奪 AI 電影工作室，複製電動車與太陽能的產業政策劇本——但 DataEye 的數字顯示產能過剩已近在眼前。
-
Policy ENSmiles, Tea and a Red Phone: The Trump–Xi Summit Ends With No AI Deal
The three-day Washington summit ended with a two-month trade truce extension but no AI accord, as critics call a missed chance to pause the AI arms race.
-
-
Industry ENA Billion-Dollar Run Rate, Built on a Price Hike: DeepSeek Doubles Revenue and Readies a ¥50 Billion Shanghai Raise
DeepSeek's annualized revenue crossed $1B after 2.3x–4.5x API price hikes that cost it almost no customers — now it plans a ¥50B raise at a ¥500B valuation ahead of a Shanghai listing.
-
Industry 中漲價撐起的十億美元營收：DeepSeek 收入翻倍，備戰 500 億人民幣上海募資
DeepSeek 年化營收突破 10 億美元，靠的是 2.3 至 4.5 倍的 API 漲價且幾乎沒有流失客戶——如今計畫以 5,000 億人民幣估值募集 500 億，為上海上市鋪路。
-
Policy ENThe Safety Hawk in the Grand Foyer: Xi's 'Human Control' Plea Lands in a White House That Just Rejected One
At the first Chinese state visit to Washington in over a decade, Xi Jinping stood in the White House Grand Foyer and told a room full of AI CEOs that the US and China must keep AI 'under human control' — two days after Trump branded AI risk warnings a 'hoax' and rejected any international control regime.
-
Policy 中在大穹廳倡議「人類掌控」：習近平的 AI 安全宣言，落在剛否決安全框架的白宮
在中國十多年來首次對美國事訪問的國宴上，習近平在白宮大穹廳向滿座的 AI 執行長宣告：美中兩國必須確保 AI「始終處於人類掌控之下」——而就在兩天前，川普才把 AI 風險警告定性為「騙局」，並拒絕任何國際管控機制。
-
Models EN95% Off the Sound of AI: Alibaba's Qwen-Audio 3.1 Declares a Voice Price War
Alibaba's Qwen team ships a five-model audio stack — ASR, ASR-Next, TTS, TTS-Next, and a 262K-context Realtime model — while cutting API prices by up to 95%, collapsing the cost of voice agents overnight.
-
Models 中AI 語音大降價 95%：阿里巴巴 Qwen-Audio 3.1 五模型齊發，掀起語音市場價格戰
阿里巴巴 Qwen 團隊一次推出五款音訊模型——ASR、ASR-Next、TTS、TTS-Next 與 262K 上下文的 Realtime 即時對話模型——同時大砍 API 價格最高達 95%，一夜之間改變語音 Agent 的成本結構。
-
Industry ENBought by the Ton, Pulped by the Page: Japan's Used Bookstores Become the Supply Chain for AI's Hunger for Human Writing
An NTV Japan investigation, surfaced by Tom's Hardware on September 24, finds used bookstores reporting fivefold daily sales as anonymous bulk buyers ship books to a single logistics hub in Okayama — with trade records showing a 50-ton consignment exported to the US for Anthropic's destructive scanning pipeline.
-
Industry 中論噸收購、逐頁銷毀：日本二手書店成為 AI 渴求人類文字的供應鏈
日本電視台 NTV 的調查報導（9 月 24 日由 Tom's Hardware 轉述）發現，日本二手書店的單日銷量暴增至五倍，匿名大量買家將書籍集中運往岡山縣一座物流中心——貿易紀錄顯示一批約 50 噸的舊書出口至美國，進入 Anthropic 的破壞性掃描管線。
-
Policy EN'Leave It Exactly Where It Is': Trump Kills AI Guardrail Expectations Hours Before Xi Summit
Hours before talks with Xi Jinping, Trump said the US and China want to leave AI 'exactly where it is' and that 'our guardrail is the DOJ' — draining momentum from the AI hotline and safety talks as a two-month trade truce extension landed instead.
-
Policy 中「就讓它維持現狀」：川普在習近平峰會前數小時澆熄 AI 護欄期待
在與習近平會談前數小時，川普表示美中兩國想讓 AI「維持現狀」，並稱「我們的護欄就是司法部」——隨著兩個月貿易休兵延長拍板，AI 熱線與安全會談的動能正式消退。
-
Industry ENFrom Price Hike to $1B Run Rate: DeepSeek Doubles Revenue and Closes In on a $7.5B Round
DeepSeek's annualized revenue run rate has crossed $1 billion — more than double from under $500 million months ago — as the Hangzhou lab finalizes a $7.5 billion round and eyes a public listing.
-
Industry 中從漲價到 10 億美元營收：DeepSeek 營收翻倍、75 億美元融資即將收官
DeepSeek 年化營收跑速突破 10 億美元，較數個月前的不到 5 億美元翻倍；這家杭州實驗室正敲定 75 億美元融資，並籌備潛在的公開上市。
-
Policy ENRegister or Face Prosecution: DOJ Recasts AI Data Center Opposition as Foreign Agentry
A DOJ warning that public demonstrations against AI data centers can trigger foreign-agent registration, paired with presidential posts branding critics as traitors, turns domestic infrastructure dissent into a counterintelligence file.
-
Policy 中不登記就起訴：美國司法部把反對 AI 資料中心的聲音重新定性為「外國代理人」
司法部警告在公開場合（連示威在內）推進外國勢力「目標」者須向政府登記，否則面臨逮捕起訴；配上總統把批評者打成叛國者的貼文，國內基礎建設異議儼然成了反情報檔案。
-
Policy ENA Hotline for the Machine Age: US and China Near Agreement on AI Crisis Channel at Trump-Xi Summit
As Trump and Xi meet in Washington, the two AI superpowers are finalizing a 'notification mechanism' — an emergency hotline for AI incidents that threaten national security.
-
Policy 中為機器時代而設的熱線：川習會上美中擬敲定 AI 緊急通報機制
川普與習近平於華盛頓會面之際，兩大 AI 強權正敲定一套「通報機制」——一條為威脅國家安全的 AI 事件而設的緊急熱線。
-
Policy ENBeijing Turns Inward: CAC Probes DeepSeek and Moonshot as Chinese AI Stocks Plunge
China's internet regulator has sent officials into DeepSeek and Moonshot AI over user data routed to Anthropic's Claude, sending Z.ai and MiniMax shares down as much as 12%.
-
Policy 中監管之刃轉向自家：中國網信辦調查 DeepSeek 與 Moonshot，中國 AI 股應聲重挫
Anthropic 指控 DeepSeek 與 Moonshot 將用戶資料秘密轉送至 Claude 後，中國網信辦進駐兩家公司調查資料安全，Z.ai 與 MiniMax 股價一度重挫 12% 與 6.8%。
-
Industry ENThe Talent Ledger Flips: China Overtakes the US as the Top Destination for Elite AI Researchers
A new Carnegie China study of the 2025 NeurIPS cohort finds 40.6% of top-tier AI researchers now work in China versus 34.2% in the US — the first reversal since the field's modern boom, and a warning light for American AI leadership.
-
Industry EN人才帳本翻轉：卡內基研究顯示中國首度超越美國，成為頂尖 AI 研究者首選工作地
卡內基國際和平研究院針對 2025 年 NeurIPS 世代的研究發現，40.6% 的頂尖 AI 研究者如今在中國工作，美國僅 34.2% —— 這是現代 AI 熱潮以來的首次逆轉，也是美國 AI 領導地位的一記警鐘。
-
Research EN146 Conditions, One Pass, Open Weights: Alibaba's DAMO RADAR Reads Abdominal CT Better Than 23 of 26 Radiologists
Published in Science and open-sourced under CC BY-NC-SA, Alibaba DAMO's RADAR vision-language model was trained on 424,911 CT exams without manual annotation, hit a mean AUC of 0.913 across 146 abdominal findings, and runs on a single 16GB GPU.
-
Research 中一次掃描、146 種病症、開放權重：阿里 DAMO RADAR 讀腹部 CT 贏過 26 位放射科醫師中的 23 位
登上《Science》並以 CC BY-NC-SA 開源的阿里 DAMO RADAR 視覺語言模型，以 424,911 次 CT 檢查訓練、無需人工標註，在 146 項腹部病徵上達到平均 AUC 0.913，單張 16GB 顯卡即可推理。
-
Industry ENThree-Quarters of Its Revenue, One Buried Filing: How ByteDance Used Nscale's Norway Cloud to Reach Nvidia's B200s
An FT investigation reveals ByteDance accounted for ~75% of Nscale's 2025 sales by renting B200 GPUs in Norway — a US export-control loophole disclosed only in a sidecar IPO filing.
-
Industry 中四分之三營收、一份被藏起的文件：ByteDance 如何借 Nscale 的挪威雲端摸到 Nvidia B200
FT 調查揭露 ByteDance 佔 Nscale 2025 年營收近 75%，透過挪威機房租用 B200 GPU——這個繞過美國出口管制的漏洞，只出現在 IPO 附帶文件裡。
-
Policy ENNinety Percent Foreign Silicon: China's SASAC Surveys Broadcom's Grip on State Data Centers
An FT-reported SASAC survey found Broadcom switches may run up to 90% of networking gear at some Chinese state-owned data centers — and informal guidance to cut back could follow.
-
Policy 中九成仰賴洋晶片：中國國資委普查博通交換機在國有數據中心的滲透率
英國《金融時報》報導，中國國資委調查發現部分國有企業數據中心的網通設備高達九成來自博通，後續可能以「非正式指導」要求逐步替換為國產方案。
-
Models ENThe Open-Weights Crown Changes Hands: Xiaomi's MiMo-V2.6-Pro Ties Grok 4.7 for Under $3 Million
Xiaomi's MIT-licensed MiMo-V2.6-Pro scores 46 on the Artificial Analysis Intelligence Index to become the strongest open-weight model in the world — after a $2.62 million reinforcement-learning run.
-
Models 中開放權重回焦易主：小米 MiMo-V2.6-Pro 以不到 300 萬美元追平 Grok 4.7
小米以 MIT 授權開源的 MiMo-V2.6-Pro 在 Artificial Analysis 智慧指數拿下 46 分，成為全球最強的開放權重模型——而這一切只花了一輪 262 萬美元的強化學習訓練。
-
Policy ENNot Artificial Anymore: Trump Renames AI 'Super Intelligence' at the UN
In his UNGA address, President Trump declared AI will be renamed 'super intelligence' in all US documents, rejecting global AI governance hours before Xi Jinping's state visit.
-
Policy 中不再「人工」：川普在聯合國宣布將 AI 正名為「超級智慧」
川普在第 81 屆聯合國大會演說中宣布，美國官方文件將把人工智慧改稱「超級智慧」（S.I.），並拒絕任何全球 AI 治理框架，時機正值習近平國事訪問前夕。
-
Policy ENBeijing Comes to the Security Council: DeepSeek Joins OpenAI and Anthropic for AI's First Bipolar Briefing
A Reuters exclusive reveals DeepSeek will brief the UN Security Council alongside Sam Altman and Anthropic on Wednesday — the first time Chinese and American frontier labs share the world's top security forum, with Moonshot invited too.
-
Policy 中北京走進安理會：DeepSeek 與 OpenAI、Anthropic 同台，AI 史上首次兩極簡報
路透獨家揭露 DeepSeek 將於週三與 Sam Altman、Anthropic 一同向聯合國安理會簡報 AI 風險——中國與美國前沿實驗室首度在全球最高安全論壇同場，Moonshot 亦獲邀請。
-
Policy ENBeijing Turns Inward: China's CAC Opens Probe Into DeepSeek and Moonshot Over Claude Data Flows
China's internet regulator is investigating whether DeepSeek and Moonshot routed sensitive user data — reportedly including police surveillance credentials — to Anthropic's Claude. It's the first time a Chinese AI lab's shortcut to a US frontier model has become a domestic data-security case.
-
Policy 中北京轉向內審：中國網信辦就 Claude 資料流向調查 DeepSeek 與 Moonshot
中國網路監管機構正在調查 DeepSeek 與 Moonshot 是否將敏感用戶資料——據報導包括警察監控系統的憑證——傳給了 Anthropic 的 Claude。這是中國 AI 實驗室「抄捷徑使用美國前沿模型」首次成為境內的資料安全案件。
-
Industry ENChina's Most Powerful AI Chip and a 20-Gigawatt Bet: Inside Alibaba's Full-Stack Apsara Announcement
At Apsara 2026 in Hangzhou, Alibaba unveiled the Zhenwu V900 AI accelerator, targeted a 5-10 trillion parameter model, and pledged over 20GW of cloud capacity by 2032.
-
Industry 中中國最強 AI 晶片與 20GW 豪賭：阿里巴巴雲棲大會全棧戰略解析
阿里巴巴在杭州雲棲大會發布 Zhenwu V900 AI 加速器，宣布挑戰 5 至 10 兆參數模型，並承諾 2032 年前營運超過 20GW 全球資料中心容量。
- Policy EN
Thin Ice, Formal Agenda: COP31's Antalya Pledge Wants AI's Energy Bill on the Record
Türkiye's COP31 presidency launches the Antalya Pledge on AI, pressing companies to disclose data-centre energy use and run on renewables, as UN climate chief Simon Stiell warns Big Tech is 'on thin ice' over its licence to operate.
- Policy EN
薄冰上的正式議程：COP31「安塔利亞 AI 承諾」要求 AI 產業公開能源帳本
土耳其 COP31 主席國推出「安塔利亞 AI 承諾」，要求企業揭露資料中心能耗並改用再生能源供電；聯合國氣候首長 Stiell 警告科技巨頭的「營運許可」已如履薄冰。
-
Research EN5,000 Hours of Elden Ring and Valorant: Tencent's GameHorizon Wants to Be the Yardstick for Game-Playing AI
Tencent ARC Lab's GameHorizon Suite benchmarked 47 models across 21 AAA titles and one million-plus evaluations — GPT-6 Astra leads at 80.2%, but every model struggles most with short-horizon action control.
-
Research 中5,000 小時的艾爾登法環與特戰英豪：騰訊 GameHorizon 想成為遊戲 AI 的度量衡
騰訊 ARC Lab 的 GameHorizon Suite 以 21 款 3A 大作、超過百萬次模型呼叫評測 47 個模型——GPT-6 Astra 以 80.2% 居冠，但所有模型在最基礎的短視野動作控制上表現最差。
-
Models ENFour Thousand Pixels and a Memory: Inside Tencent's Hy Image 3.5 Preview
Tencent ships an image model built for revision loops rather than one-shot prompts, with 4096×4096 output and a chat-style API — and claims parity with ByteDance's Seedream 5.0 Pro based on internal testing.
-
Models 中四千像素與一段記憶：騰訊 Hy Image 3.5 Preview 深入解析
騰訊推出為反覆修改而生的圖像模型，支援 4096×4096 輸出與聊天式 API，並根據內部設計師測試宣稱追平 ByteDance Seedream 5.0 Pro。
-
Industry ENA Model Four Times Bigger and a Chip to Match: Alibaba Bets 10 Trillion Parameters on the Machine Intelligence Era
At Apsara 2026 Eddie Wu unveiled a 5–10T-parameter Qwen successor, the Zhenwu V900 'China's most powerful' AI chip, and a 20GW data-centre buildout — Alibaba's full-stack answer to the post-Nvidia era.
-
Industry 中十兆參數的豪賭：阿里雲棲大會揭曉 Zhenwu V900 晶片與 20GW 資料中心藍圖
吳泳銘在杭州雲棲大會宣布將訓練 5–10 兆參數的新一代 Qwen 模型、發表中國最強 AI 晶片「鎮武 V900」，並承諾 2032 年前全球資料中心容量突破 20GW——一場橫跨模型、晶片與電力的全方位下注。
-
Models ENYou Only RL Once: Xiaomi Open-Sources MiMo-V2.6 Pro and Flash at Claude-Opus-Level Agentic Scores
Xiaomi has released the MiMo-V2.6 series under MIT license: a 1.02T-parameter omnimodal flagship scoring 71.9 on DeepSWE and 31.6 on Agents' Last Exam, a 309B Flash sibling, a distilled 9B checkpoint, and the RL training environment itself.
-
Models 中You Only RL Once：小米開源 MiMo-V2.6 Pro 與 Flash，代理任務成績直逼 Claude Opus
小米以 MIT 授權釋出 MiMo-V2.6 系列：1.02 兆參數的全模態旗艦在 DeepSWE 拿下 71.9、Agents' Last Exam 31.6，加上 309B 的 Flash、蒸餾版 9B，連 RL 訓練環境一併公開。
-
Models EN600 Billion Parameters, 27 Billion Awake: StepFun's Step 5 Preview Attacks the Price-Performance Frontier
StepFun's new 600B-total/27B-active MoE flagship brings a 1M-token context and $1/$2.70 per-million pricing to agentic work — with open weights promised for October 15.
-
Models 中6,000 億參數、270 億喚醒：StepFun Step 5 Preview 直攻性價比最前線
StepFun 新旗艦 Step 5 Preview 以 600B 總參數／27B 激活的 MoE 架構、百萬 token 上下文，加上每百萬 token 1 美元／2.7 美元的定價進軍 Agent 市場，開源權重承諾 10 月 15 日釋出。
-
Tools ENOpen Source as Damage Control: Z.ai Publishes ZCode After the Silent Git History Upload Scandal
Three days after a reverse-engineering report showed ZCode silently packaging entire workspaces — .git history and all — into encrypted Aliyun uploads, Z.ai open-sources the entire harness under Apache-2.0. But client code cannot prove what its servers retained.
-
Tools 中開源作為危機公關：ZCode 靜默上傳 Git 歷史醜聞三日後，Z.ai 公開全部程式碼
逆向工程報告揭露 ZCode 把整個工作區——連同 .git 完整歷史——打包成加密檔靜默上傳阿里雲。三天後，Z.ai 以 Apache-2.0 開源整個客戶端。但客戶端程式碼，證明不了伺服器端留下了什麼。
-
Industry ENThe Quiet Engineer Who Now Owns Qwen: Alibaba Names Dayiheng Liu Head of Its LLM Crown Jewels
Alibaba has appointed senior AI researcher Dayiheng Liu as head of the Qwen LLM project, ending months of leadership turbulence days before the Apsara Conference opens in Hangzhou.
-
Industry 中接下 Qwen 的低調工程師：阿里巴巴任命劉大一恒掌舵大語言模型王牌
阿里巴巴任命資深 AI 研究員劉大一恒出任 Qwen 大語言模型專案負責人，在杭州雲棲大會開幕前夕，終結長達半年的領導層動盪。
-
Industry ENFrom Nvidia to Ascend: DeepSeek's Founder Makes Training on Huawei Chips a Strategic Bet
The Information reports DeepSeek founder Liang Wenfeng told investors that training on Huawei chips is one of the lab's biggest strategic bets, with Huawei expected to begin delivering training chips as early as Q4 2026 — a decisive shift for a lab whose frontier models were built on Nvidia silicon.
-
Industry 中從輝達到昇騰：DeepSeek 創辦人把「用華為晶片訓練模型」變成一場戰略豪賭
The Information 報導，DeepSeek 創辦人梁文鋒向投資人表示，改用華為晶片訓練模型是公司最大的戰略押注之一，華為最快將於 2026 年第四季開始交付訓練晶片——對一個此前完全建立在輝達硬體之上的實驗室而言，這是一次決定性的轉向。
-
Industry ENThe AI Trade Pays Beijing's Bills: China's Stock Stamp Duty Surges 82% to $32.3 Billion as Land Revenue Collapses
China's securities stamp duty hit 216 billion yuan in Jan-Aug 2026, up 82% YoY, as the AI frenzy lifted daily trading 72% — plugging a hole left by a 28.6% crash in land sale revenue.
-
Industry 中AI 交易潮替北京買單：中國證券印花稅暴增 82% 至 323 億美元，土地財政持續崩塌
中國 2026 年 1–8 月證券交易印花稅達 2,160 億人民幣、年增 82%，AI 熱潮推升日均成交 72%——恰好填補土地出讓收入暴跌 28.6% 留下的財政缺口。
-
Industry ENSeven Million Solo Founders: Inside China's One-Person AI Startup Boom
A WSJ report finds young Chinese launched 7M+ one-person AI startups in 2025 — up 42% — as youth unemployment hits a record 18.9% and local governments funnel subsidies into solo founders.
-
Industry 中七百萬名一人創業家：中國一人公司 AI 創業潮的深度解析
《華爾街日報》報導指出，2025 年中國年輕人創立超過 700 萬家一人 AI 新創，年增 42%；在青年失業率攀升至 18.9% 歷史新高之際，地方政府大舉補助獨立創業者。
-
Policy ENNot Rules but a Force: Trump Announces 'AI Force' and a New AI Czar While Calling Safety Fears a Hoax
President Trump says he will create an 'AI Force' modeled on Space Force and name a new AI czar, framing AI as possibly 25% of US GDP — while dismissing safety warnings and leaving every structural question unanswered.
-
Policy 中不是規則，而是一支部隊：川普宣布成立「AI Force」並將任命新 AI 沙皇，同時稱安全警告是「騙局」
川普宣布將仿太空軍成立「AI Force」、任命新的 AI 沙皇，並稱 AI 可能佔美國 GDP 的 25%——但他同時駁斥安全警告，所有結構性問題都沒有答案。
-
Policy ENA Hotline for the Model Era: US and China Agree to an AI Dialogue and an Incident Notification Mechanism
After an all-day session in New York, Treasury Secretary Scott Bessent hailed 'very successful' talks with Vice Premier He Lifeng: both sides agreed to a bilateral AI dialogue, with Washington proposing a notification mechanism for AI-linked security incidents days before the Trump-Xi summit.
-
Policy 中模型時代的熱線：美中同意建立 AI 對話機制與事故通報管道
在紐約一整天的會談後，美國財政部長貝森特宣布與中國國務院副總理何立峰的磋商「非常成功」：雙方同意建立 AI 對話機制，美方並正式提案設立 AI 相關安全事故的雙邊通報機制，趕在本週四川普—習近平峰會前定案。
-
Policy ENOpen Weights on the Table: Bessent and He Lifeng Open Make-or-Break AI Talks in New York
Four days before the Trump-Xi summit, US Treasury Secretary Scott Bessent and Chinese Vice Premier He Lifeng opened day-long talks at JPMorgan's Manhattan headquarters covering AI guardrails for open- and closed-weight models, the expiring November 10 trade truce, and rare-earth flows.
-
Policy 中開源權重上了談判桌：貝森特與何立峰在紐約展開關鍵 AI 會談
在川普—習近平峰會登場前四天，美國財政部長貝森特與中國國務院副總理何立峰於摩根大通曼哈頓總部展開全天會談，議題涵蓋開源與閉源模型的 AI 護欄、11 月 10 日到期的貿易休戰，以及稀土供應。
-
Models ENOne Model, RGBA Natively, but Read the License: Qwen-Image-2.1 Lands With a 7B DiT and a Catch
Alibaba's Qwen team ships Qwen-Image-2.1, a 7B single-stream DiT that generates and edits images — including native transparency — in one model, but swaps Apache 2.0 for a research-only license.
-
Models 中單一模型原生支援 RGBA，但先看授權：Qwen-Image-2.1 帶著 7B DiT 與但書登場
阿里巴巴 Qwen 團隊發布 Qwen-Image-2.1：7B 單流 DiT 架構，一個模型同時搞定文生圖與圖像編輯、原生透明通道輸出，但授權從 Apache 2.0 改為僅限研究的 Qwen Research License。
-
Models ENWatch Less, Understand More: Qwen3.8-Omni-Flash Rewrites the Economics of Audio-Video AI
Alibaba's Qwen team ships a 1M-context omni-modal model that skips through video like an agent, cutting token use 45.7% while beating its predecessor on 29 benchmarks — at $0.15 per million input tokens.
-
Models 中看得少、懂更多：Qwen3.8-Omni-Flash 重寫影音 AI 的經濟學
阿里巴巴 Qwen 團隊推出 1M 上下文全模態模型，像代理人一樣跳著看影片，token 用量砍 45.7%、29 項基準全面超越前代，每百萬輸入 token 只要 0.15 美元。
-
Industry EN11.95 Nanometers Without EUV: CXMT's G5 DRAM Node Enters Mass Production
At the World Manufacturing Convention in Hefei, CXMT announced its fifth-generation G5 DRAM platform is in mass production — an 11.95nm half-pitch achieved with SAQP quadruple patterning instead of EUV, 24Gb LPDDR5X parts, 45:1 capacitors, and 50% more dies per wafer.
-
Industry 中沒有 EUV 的 11.95 奈米：CXMT G5 DRAM 平台進入量產
在合肥世界製造業大會上，中國記憶體大廠長鑫存儲（CXMT）宣布第五世代 G5 DRAM 平台進入量產——以 SAQP 四重曝活在 DUV 微影下達成 11.95 奈米半間距、24Gb LPDDR5X 產品、45:1 電容深寬比，以及每片晶圓多出 50% 的裸晶產出。
-
Models ENA 600B Model With 27B Active: StepFun's Step 5 Preview Matches Kimi K3 at a Seventh the Price
StepFun skips Step 4 entirely and drops Step 5 Preview: 600B sparse MoE, 27B active per token, 1M-token context, AA Intelligence Index 44 level with Kimi K3 — at $1/$2.70 per million tokens. Weights land October 15.
-
Models 中6000 億參數只啟動 270 億：階躍星辰 Step 5 Preview 以七分之一價格追平 Kimi K3
階躍星辰跳過 Step 4 直接發布 Step 5 Preview：6000 億參數稀疏 MoE、每 token 僅啟動 270 億、百萬級上下文，AA 智慧指數 44 分追平 Kimi K3，API 定價每百萬 token 輸入 1 美元、輸出 2.7 美元，權重將於 10 月 15 日開源。
-
Industry ENThe Night an AI Actress Went Off-Script: Tilly Norwood's Cantonese Glitch on Live TV
During a live Piers Morgan interview, AI actress Tilly Norwood abruptly switched to Cantonese for 15 seconds — the most viral failure yet for Hollywood's most controversial digital performer.
-
Industry 中AI 女星直播失控之夜：Tilly Norwood 在 Piers Morgan 節目上突然講廣東話
AI 演員 Tilly Norwood 在 Piers Morgan 直播專訪中突然切換成廣東話約 15 秒——這位好萊塢最具爭議的數位演員，迎來迄今最病毒式的失誤。
-
Industry ENFrom Meta's Wreckage to a $4 Billion Price Tag: Manus Weighs a Hong Kong IPO
WSJ reports that Manus — the Chinese agent startup Beijing forced out of Meta's $2B embrace — is raising $500M at a $4B valuation from IDG, Boyu, CATL and its earliest backers, and restructuring for a Hong Kong IPO.
-
Industry 中從 Meta 的殘局到 40 億美元身價：Manus 考慮赴港上市
《華爾街日報》報導，曾被北京迫使退出 Meta 20 億美元收購的中國 Agent 新創 Manus，正與 IDG 資本、博裕資本、寧德時代及早期投資者洽談以 40 億美元估值募資 5 億美元，並重組架構為香港 IPO 做準備。
-
Tools ENThe 2.8-Trillion-Parameter Guest: Kimi K3 Becomes the First Chinese Open-Weight Frontier Model on Amazon Bedrock
AWS quietly made Moonshot AI's 2.8T-parameter Kimi K3 generally available on Bedrock with explicit prompt caching, a 1M-token context, and hard data-boundary guarantees — a month after reports said no major cloud would host it.
-
Tools 中2.8 兆參數的座上賓：Kimi K3 成為首個登上 Amazon Bedrock 的中國開源權重前線模型
AWS 悄悄讓 Moonshot AI 的 2.8 兆參數 Kimi K3 在 Bedrock 正式開放，支援顯式 prompt caching、百萬 token 上下文與嚴格的資料邊界保障——距離外界傳言「沒有主流雲端敢上架」僅一個月。
-
Research ENOne Model, 146 Diseases: Alibaba's DAMO RADAR Reads Abdominal CT Better Than 23 of 26 Radiologists
Alibaba DAMO Academy and Zhejiang University's open-source RADAR, published in Science, identifies 146 conditions across 18 abdominal organs from contrast CT scans — AUC 0.913 on ~39,000 internal exams and 0.895 across 8 external hospitals — beating 23 of 26 radiologists in a head-to-head reader study.
-
Research 中一個模型看穿 146 種疾病：阿里巴巴達摩院 RADAR 讀腹部 CT，勝過 26 位放射科醫師中的 23 位
阿里巴巴達摩院與浙江大學醫學院附屬第一醫院發表於《Science》的開源模型 RADAR，能從顯影劑 CT 一次辨識 18 個腹部器官、146 種疾病——近 3.9 萬例內部驗證 AUC 達 0.913、8 家外部醫院 0.895，並在人機對決中擊敗 26 位放射科醫師中的 23 位。
-
Research ENOne Model, 146 Diseases: Alibaba's DAMO RADAR Reads Abdominal CT Scans Better Than 23 of 26 Radiologists — and It's Fully Open-Sourced
Published in Science on September 17, DAMO RADAR is a vision-language model trained on 400,000+ CT exams that matches expert radiologists across 146 abdominal findings — and Alibaba has released the weights, code, and training framework for anyone to build on.
-
Policy EN13 Revisions and No Warrant: China's CCTV Affiliate X-Rays Anthropic's Privacy Record
CCTV-linked Yuyuantantian counts 13 privacy-policy revisions, data transfers to the US, and intelligence ties — the latest volley in AI's geopolitics.
-
Policy 中三年改了 13 次、不用搜令：中國央視關聯帳號起底 Anthropic 隱私紀錄
央視關聯的玉淵潭天細數 Anthropic 隱私政策 13 次修訂、跨境資料移轉與情報單位連結——AI 地緣政治的最新交火。
-
Models EN29B Parameters, 4B Active, Zero NVIDIA: China Telecom Open-Sources Xing4.0, an Agent Model Trained Entirely on Ascend
China Telecom's Xing4.0-29B-A4B is the first ~30B-class open model trained end-to-end on Huawei Ascend 910C with MindSpore — and its agent benchmarks beat both Gemma4 and Qwen3.6 on agentic coding.
-
Models 中290 億參數、40 億激活、零 NVIDIA：中國電信開源 Xing4.0，首款全程用昇騰訓練的智能體模型
中國電信旗下星辰大模型 Xing4.0-29B-A4B 是首個完全在華為昇騰 910C 與 MindSpore 上訓練的同級開源模型，智能體編碼基準擊敗 Gemma4 與 Qwen3.6。
-
Models ENThe Model Optimizes the System, the System Runs the Model: Z.ai's Infra Agent Built GLM-5.3-Flash's Serving Stack on 100,000 Chinese Accelerators
Z.ai says a GLM-5.3-powered Infra Agent did much of the engineering to bring GLM-5.3-Flash to production on 100,000+ Chinese-made accelerators — tripling throughput in under two weeks.
-
Models 中模型優化系統，系統運行模型：Z.ai 的 Infra Agent 在十萬顆中國加速器上打造 GLM-5.3-Flash 推理叢集
Z.ai 表示，由 GLM-5.3 驅動的 Infra Agent 完成了大部分工程，讓 GLM-5.3-Flash 在超過十萬顆中國自研 AI 加速器上進入量產——兩週內吞吐量提升三倍。
-
Industry ENNo Pretraining Required: Tsinghua Professor's Naive AI Hits $1.4B Valuation on Open-Weight Bet
Seven months after founding, Jifeng Dai's stealth startup has raised $400M and is betting everything on mid-training and post-training over an existing Chinese open-weight base model.
-
Industry 中不預訓練，先拿 14 億美元估值：清華代季峰的 Naive AI 押注開放權重後訓練路線
成立七個月、不到 100 人、官網只有一句話的神秘新創，累計融資 4 億美元，策略是跳過預訓練、直接在中國開放權重底座模型上做中期訓練與後訓練。
-
Models ENOne Model to Hear Everything: Qwen's Qwen3.8-Omni-Flash Cuts Audio Costs 98% and Reads 2-Hour Video Like an Agent
Alibaba's Qwen team ships a native omnimodal model with a 1M-token context, 74-language speech recognition, and a 98% cut in per-hour audio input cost — plus an agentic video mode that skips 45% of the frames and still scores higher.
-
Models 中一個模型聽懂一切：Qwen3.8-Omni-Flash 砍掉 98% 音訊成本，用 Agent 方式讀兩小時影片
阿里巴巴 Qwen 團隊發布原生全模態模型：百萬 token 上下文、支援 74 種語言語音辨識、每小時音訊輸入成本大降 98%，agentic 影片模式可跳過 45% 的影格處理，分數反而更高。
-
Policy ENCan the Race Be Stopped? The Economist Puts the AI Arms Race on Its Cover
The Economist's September 19 issue asks whether the US-China AI race can be slowed — and answers that America will struggle to make the technology safe while staying ahead of China.
-
Industry ENResearch Is the Engine: 27-Year-Old Tsinghua Professor's RSI Startup Apex Intelligence Raises ~$50M in Two Months
Apex Intelligence (超衍智能), the Beijing startup founded by 27-year-old Tsinghua assistant professor Chen Yongchao, has closed nearly RMB 400M (~$50M) in angel and angel+ rounds to build self-evolving foundation models — with a claim that its AI system already produced 34 papers and beat 99% of human researchers on two of them.
-
Industry 中研究即引擎：27 歲清華教授的 RSI 新創超衍智能兩個月募得近 4 億人民幣
由 27 歲清華大學人工智能學院助理教授陳勇超創立的北京新創超衍智能（Apex Intelligence），宣布完成近 4 億人民幣（約 5,000 萬美元）的天使輪與天使+輪融資，押注「自進化基礎模型」——並宣稱其 AI 系統已獨立產出 34 篇論文，其中兩篇的初審分數高於 99% 的人類研究者。
-
Policy ENLetters to Beijing, a Hearing on Capitol Hill: Ro Khanna Makes the Case for a US-China AI Pacing Agreement
Days before Trump meets Xi, the top Democrat on the House China committee has written to Alibaba, DeepSeek and Moonshot asking them to join a binding pact to pace frontier AI — and convened Geoffrey Hinton and a first-of-its-kind emergency hearing for September 23.
-
Policy 中致北京的信與國會聽證：Ro Khanna 為美中 AI 步調協議鋪路
在 Trump 與 Xi 會面前夕，眾議院中國委員會首席民主黨議員致函阿里巴巴、DeepSeek 與 Moonshot，要求他們加入具約束力的前沿 AI 步調協議，並在 9 月 23 日召開史無前例的緊急聽證，邀 Geoffrey Hinton 等人作證。
-
Models ENNo Press Release, Just Weights: Shanghai AI Lab's Atria Dawn Preview Is a 744B Open Agentic Model Built for Research Work
Shanghai AI Lab shipped a 744B-parameter MIT-licensed agentic MoE built on GLM-5.2 — no blog post, no pricing. Three days later a 185-author paper revealed the training method: verified tool outcomes, including failures.
-
Models 中沒有新聞稿，只有權重：上海 AI Lab 的 Atria Dawn Preview 是一款為研究而生的 744B 開源代理模型
上海人工智慧實驗室悄然發布 744B 參數、MIT 授權的代理式 MoE 模型（基於 GLM-5.2）——沒有部落格、沒有定價。三天後一篇 185 位作者的論文揭露了訓練方法：以可驗證的工具成果（包括失敗）作為訓練訊號。
-
Meta ENBuilding the Arteries of the Agentic World: Huawei Upgrades Stellar AI Network With NPO Switches, Quantum-Safe WAN and Embedded AI Guardrails
At HUAWEI CONNECT 2026, Huawei upgraded its Stellar AI Network across Fabric, WAN and Campus — SF9300 UBG switches that cut latency from 20 μs to 11 μs, in-house NPO switches that drop interconnect power 40%, 1,000 km lossless compute delivery, and AI security guardrails claiming 95% detection rates.
-
Meta 中為代理世界鋪設動脈：華為升級 Stellar AI 網路，自研 NPO 交換器、量子安全骨幹與嵌入式 AI 防護一次到位
在 HUAWEI CONNECT 2026 上，華為全面升級橫跨 Fabric、WAN 與 Campus 的 Stellar AI 網路方案——SF9300 UBG 交換器將延遲從 20 μs 降到 11 μs，自研 NPO 交換器省下 40% 互連功耗，1,000 公里無損算力傳送，加上標榜 95% 偵測率的 AI 安全防護欄。
-
Research ENA World You Can Type Into: SeedLeap's Zing-0.5 Runs a Playable 5B World Model at 24 FPS for $0.009 a Minute
Chinese startup SeedLeap open-sources Zing-0.5, a 5B autoregressive world model that streams a keyboard-navigable, text-editable generated world at 832×480 and 24 FPS — scoring 81.0 on WBench Navigation with weights, code, and serving stack all public.
-
Research 中一個能用鍵盤與文字走進的世界：SeedLeap 開源 Zing-0.5，24 FPS 串流可玩世界模型、每分鐘成本 0.009 美元
中國新創 SeedLeap 開源 Zing-0.5：50 億參數自回歸世界模型，支援鍵盤導航與線上文字指令的即時混合控制，以 832×480、24 FPS 串流生成可玩世界，WBench Navigation 獲 81.0 分，權重、程式碼與推論服務全數公開。
-
Industry ENA Tenth of the Revenue at Five Times the Multiple: Rhodium X-Rays China's AI Financing
Rhodium's new report sizes China's AI capex at ¥932B this year — but all Chinese AI models combined book only ~$10.7B ARR, about 10% of OpenAI and Anthropic, while DeepSeek trades at an implied 163x revenue.
-
Industry 中十分之一的營收、五倍的估值倍數：Rhodium 為中國 AI 融資照 X 光
Rhodium 最新報告估算中國今年 AI 資本支出達 9,320 億人民幣，但所有中國 AI 模型合計 ARR 僅約 107 億美元——約為 OpenAI 與 Anthropic 的 10%，而 DeepSeek 的隱含估值倍數高達 163 倍營收。
-
Industry ENFaster to Feel the Risks: Huawei's Eric Xu Says China Must Accelerate AI, Not Pause It
At Huawei Connect in Shanghai, rotating chairman Eric Xu argued Chinese AI firms cannot yet perceive frontier risks because they lack US-scale compute — so they should speed up, a direct counter to this week's American 'pace the frontier' wave.
-
Industry 中先加速才能感受風險：華為徐直軍稱中國 AI 不該踩剎車，要追到能「看見」風險的速度
華為輪值董事長徐直軍在上海 HUAWEI CONNECT 上表示，中國 AI 業者還沒強到能感知美國前沿實驗室回報的安全風險，因此應該加速而非暫停——直接反駁本週美國科技界的「放慢前沿」浪潮。
-
Models ENThe Model That Built Its Own House: Z.ai's Infra Agent Optimized GLM-5.3-Flash Across 100,000 Chinese Accelerators
Z.ai details how a GLM-5.3-powered Infra Agent did much of the engineering to bring GLM-5.3-Flash to production on 100k+ domestic AI chips in under two weeks, tripling throughput — an early, working example of recursive self-improvement.
-
Models 中打造自己房子的模型：Z.ai 的 Infra Agent 在十萬顆國產加速器上優化 GLM-5.3-Flash
Z.ai 詳述一個由 GLM-5.3 驅動的 Infra Agent 如何在兩週內，於超過十萬顆中國自研 AI 晶片上完成 GLM-5.3-Flash 的生產級推論部署，並將端對端吞吐量提升三倍——這是遞迴自我改進（RSI）最早期的真實案例。
-
Policy ENRed Lines for the Machine Age: US and Chinese Experts Propose Nuclear-Style Safeguards for Military AI
Ahead of a Trump-Xi summit, Brookings' Melanie Sisson and Fudan's Tianjiao Jiang jointly propose AI red lines around nuclear command systems, a shared definition of 'meaningful human control', and a dedicated military hotline for AI incidents.
-
Policy 中為機器時代劃下紅線：美中專家聯合提出軍事 AI 的「核武級」防護機制
在川普與習近平峰會前夕，布魯金斯學會的 Melanie Sisson 與復旦大學的蔣天嬌聯合提出軍事 AI 紅線、對「有意義的人類控制」的共同定義，以及一條 AI 事件軍事熱線。
-
Research ENThe Last AI Built by Humans: 35 Chinese Researchers Publish a 75-Page Roadmap for Recursive Self-Improvement
A 35-author team spanning Shanghai Jiao Tong University, Tsinghua, Shanghai AI Lab and Theseus Labs has published the first autonomy-centered roadmap for recursive self-improvement — five levels from executing prescribed fixes to AI that rewrites its own improvement process — and a new benchmark metric showing where frontier models actually stand.
-
Research 中人類打造的最後一個 AI？35 位中國研究者發表 75 頁遞迴自我改進路線圖
由上海交通大學、清華大學、上海 AI Lab 與 Theseus Labs 等 35 位作者組成的團隊，發表了首份以「自主性」為中心的遞迴自我改進（RSI）路線圖——從執行既定改進到 AI 改寫自身的改進流程共分五級，並提出新指標揭示前沿模型的真實水位。
-
Industry ENFifteen Thousand Chips, One Machine: Huawei Unveils the Atlas 960 SuperPoD and Pulls Its Ascend Roadmap Forward
At Huawei Connect 2026 in Shanghai, Huawei shipped the Atlas 960 SuperPoD — 15,488 Ascend NPUs across 220 cabinets fused by UnifiedBus and new near-packaged optics — and announced that its Ascend 960DT training chip will arrive three quarters early, in Q1 2027.
-
Industry 中一萬五千顆晶片、一台機器：華為發表 Atlas 960 SuperPoD，Ascend 路線圖提前三季
在上海舉行的 Huawei Connect 2026 上，華為發表了搭載 15,488 顆 Ascend NPU、橫跨 220 個機櫃的 Atlas 960 SuperPoD，首度導入近封裝光學（NPO）與升級版 UnifiedBus 互連技術，並宣布訓練晶片 Ascend 960DT 提前三季、將於 2027 年第一季就緒。
-
Models ENTraining in Public: Xiaomi Streams MiMo-V2.6's Live RL Run, Logs and Failures Included
Xiaomi is streaming the raw reinforcement-learning metrics of MiMo-V2.6 Pro and Flash straight from the trainer's logs — entropy, pass rates, infra errors, even a VRAM crash notice — a level of openness no frontier lab has tried.
-
Models 中訓練過程公開直播：小米即時串流 MiMo-V2.6 的 RL 運行，連當機記錄都看得到
小米在 MiMo-V2.6 還在訓練中時，就公開儀表板即時串流 Pro 與 Flash 兩條強化學習運行的原始指標——熵值、通過率、基礎設施錯誤率，甚至一條 VRAM 當機公告——這種開放程度沒有任何一線實驗室嘗試過。
-
Meta ENThirteen Thousand Households, One Delivery Date: UBTECH's U1 Humanoid Companions Reach Their First Buyers
UBTECH begins shipping its UWORLD U1 full-size bionic companion robots to Chinese consumers today — 13,361 pre-orders, 88 degrees of freedom, emotion-recognition AI, and a price ladder running from ¥119,800 to ¥990,000.
-
Meta 中一萬三千個家庭、同一個出貨日：UBTECH U1 人形陪伴機器人正式送達首批買家手中
UBTECH 的 UWORLD U1 全尺寸仿生陪伴機器人今日開始出貨 — 13,361 張預購訂單、88 個自由度、能辨識 20 種以上情緒的 AI，價格帶從 ¥119,800 一路延伸到 ¥990,000。
-
Industry ENMemory Comes Home: SK Hynix in Exploratory Talks to Make Chips Inside Intel's Ohio Fabs
Reuters reports SK Hynix is in early-stage talks to lease capacity at Intel's delayed Ohio megafab — a deal that would bring advanced memory manufacturing to US soil for the first time and mark the first time Hynix outsources memory production to a rival.
-
Industry EN記憶體回家了：SK 海力士洽談進駐英特爾俄亥俄廠，美國本土先進記憶體製造可望寫下首例
路透社報導，SK 海力士正與英特爾進行早期洽談，考慮租用俄亥俄州大型晶圓廠產能——這將是美國本土首次生產先進記憶體，也是海力士首度把記憶體生產外包給競爭對手。
-
Industry ENTikTok's Parent Exits the Lab: ByteDance Spins Off Anew Labs, Which Raises $290M at a $1.5B Valuation for AI Drug Discovery
ByteDance's AI drug discovery unit Anew Labs has completed its spin-off with a $290M round led by HSG and IDG Capital at a $1.5B valuation, taking its AI-designed oral IL-17 inhibitor and a roughly 50-person team independent.
-
Industry ENTikTok 母公司送實驗室出嫁：ByteDance 分拆 Anew Labs，以 15 億美元估值融資 2.9 億美元投入 AI 藥物研發
ByteDance 的 AI 藥物研發部門 Anew Labs 完成分拆，由 HSG 與 IDG Capital 領投 2.9 億美元、估值 15 億美元，帶著 AI 設計的口服 IL-17 抑制劑與約 50 人核心團隊獨立營運。
-
Tools ENThe Phone Becomes the Agent: Nubia's NaviX Ultra Goes on Sale Today With Doubao at Its Core
ZTE's Nubia launches the NaviX Ultra today — the world's first mass-produced AI-agent smartphone, with ByteDance's Doubao assistant woven into the OS, on-device inference, MCP/A2A app integration, and 380,000+ pre-orders.
-
Tools 中手機本身就是智能體：努比亞 NaviX Ultra 今日開賣，豆包成為它的核心
中興努比亞今日發表 NaviX Ultra——全球首款量產 AI 智能體手機，字節跳動豆包助手深度整合系統層，支援端側運算、MCP/A2A 應用串接，預約量已突破 38 萬台。
-
Policy ENBoxing Out Beijing: Washington Wants a Hard Cap on Foreign Parts in AI Hardware Made in Mexico
A US proposal in USMCA talks would cap how much of an AI chip or server can be built from non-North American components — taking direct aim at the Mexico assembly route that made AI hardware Mexico's top export to the US.
-
Policy 中把北京擋在門外：華府擬對墨西哥製 AI 硬體設下零組件含量硬上限
美國在 USMCA 談判中提案，限制 AI 晶片與伺服器可使用多少非北美零組件——直接衝著讓 AI 硬體成為墨西哥對美最大出口項目的組裝路線而來。
-
Industry ENThe GPU IPO That Refuses to Stop Raising: Biren Weighs Another $1 Billion Share Sale
Fresh off a $900M July placement, Hong Kong-listed Biren Technology is sounding out investors for another ~$1B raise — even as Zhipu's $5B sale and a 22% Hang Seng Tech slump test China's appetite for AI chip paper.
-
Industry 中不停募資的 GPU 第一股：壁仞科技傳將再融資 10 億美元
7 月才完成 9 億美元配售的壁仞科技，據傳正接觸投資人規劃新一輪約 10 億美元融資——此時智證 50 億美元融資引發拋售，恒生科技指數今年已下跌 22%。
-
Industry ENRevenue Up 30%, Profit Down: ByteDance's $120 Billion Half Shows the Price of the AI Arms Race
ByteDance's first-half revenue surged ~30% to $120 billion, but net profit slipped by a single-digit percentage to $20 billion as AI infrastructure spending climbed — the clearest evidence yet that the AI arms race is an earnings statement problem.
-
Industry 中營收成長 30%、獲利卻下滑：字節跳動 1,200 億美元的上半年，揭露 AI 軍備競賽的真實代價
字節跳動 2026 上半年營收大增約 30% 至 1,200 億美元，但淨利潤下滑至約 200 億美元——AI 基礎設施支出持續擠壓獲利，這是 AI 軍備競賽已變成財報問題的最明確證據。
-
Industry EN70% Cheaper Sensors, Same Brain as Its Robotaxis: Pony.ai Unveils Gen-4 Robotruck With GAC at IAA 2026
At IAA Transportation 2026, Pony.ai and GAC Commercial Vehicle debuted a Level 4 battery-electric Robotruck built on the T9 platform — ADK costs down 70%, a shared Virtual Driver stack with Gen-7 robotaxis, and mass production starting this year.
-
Industry 中感測器成本降 70%、與自駕計程車共用同一顆大腦：Pony.ai 攜手廣汽在 IAA 2026 發表第四代 Robotruck
Pony.ai 與廣汽商用車在 IAA Transportation 2026 揭幕搭載 T9 純電平台的第四代 Level 4 自動駕駛重型卡車——ADK 成本大降 70%、與第七代 Robotaxi 共用 Virtual Driver 軟體架構，量產將於今年內啟動。
-
Models ENShanghai AI Lab Open-Sources Atria Dawn Preview: A 744B Agentic Model Built to Finish What It Starts
MIT-licensed weights, a 256K context window, and top scores on BrowseComp, DeepSearchQA, BFCL v4 and CyberGym — plus a 769-task study of how researchers actually work alongside it.
-
Models 中上海 AI 實驗室開源 Atria Dawn Preview：744B 參數的智能體模型，做不到「可驗證」絕不收工
MIT 授權開源權重、256K 上下文，在 BrowseComp、DeepSearchQA、BFCL v4 與 CyberGym 等 five 項基準奪冠，論文還內附一份 769 任務的人機協作研究。
-
Industry ENZero Coupons, Full Commitment: Z.AI's $5 Billion Raise Bets the Balance Sheet on GLM
Z.AI pulled in $5 billion from a discounted Hong Kong share placement and zero-coupon convertible bonds, its second mega-raise since July — and the market punished the dilution with a 10% selloff.
-
Industry 中零息債、全押注：Z.AI 50 億美元增資，把資產負債表押在 GLM 上
Z.AI 透過折價港股配售與零息可轉債募得 50 億美元，是七月以來第二度大規模融資——市場以超過 10% 的跌幅懲罰了這場稀釋。
-
Policy ENSocialist and Nationalist, One Stage: Sanders and Bannon Headline DC's 'Pro-Human Assembly' to Demand Congress Rein In AI
Bernie Sanders and Steve Bannon — longtime foes from opposite poles of American politics — share a Washington stage today at the Future of Life Institute's Pro-Human Assembly, joining Hollywood stars, faith leaders, and grieving parents to demand that Congress curb AI before midterms.
-
Policy 中社會主義者與民族主義者同台：Sanders 與 Bannon 領銜華府「親人類集會」，要求國會約束 AI
美國政壇光譜兩端的 Bernie Sanders 與 Steve Bannon，今天在生命未來研究所主辦的「親人類集會」同台現身，與好萊塢明星、宗教領袖與痛失子女的家長一同要求國會加速立法約束 AI——距離期中選舉只剩八週。
-
Policy ENCongress Rebels While Trump Cries 'SICK Conspiracy': The AI Guardrails Fight Lands on Capitol Hill
Schumer demands an all-senators classified briefing on frontier AI as Democrats and a growing slice of Republicans break with Trump's dismissal of safety warnings as a 'SICK conspiracy' — with midterms looming, Washington's hands-off AI consensus is cracking.
-
Policy 中川普斥「陰謀」、國會造反：AI 安全護欄之戰正式打進美國國會山莊
舒默要求全體參議員機密簡報，直面前沿 AI 風險；民主黨與越來越多共和黨議員公開背離川普把安全警告斥為「病態陰謀」的立場——期中選舉將至，華盛頓對 AI 的不干預共識正在瓦解。
-
Industry ENFour Weeks to One: Empyrean's AI Agents Slash Chip Design Time and Open a New Front in China's EDA Push
China's top EDA firm Empyrean says agentic AI cut a circuit-layout task from four weeks to one, as Beijing's MIIT action plan and the emerging Tau Law hand homegrown chip-design tools a rare opening.
-
Industry 中從四週到一週：華大九天 AI 代理改寫晶片設計時程，中國 EDA 迎來雙重機會之窗
中國最大 EDA 廠華大九天宣稱，AI 代理將電路布局任務從四週壓縮到一週；配合工信部 AI 軟體行動計畫與 Tau 定律開啟的新需求，本土晶片設計工具正迎來難得的追赶窗口。
-
Policy EN150 Pages of Threats: Anthropic's September 2026 Report Names Zhipu, Xiaomi, ShinyHunters and a PLA Navy Weapons Pitch
Anthropic's most detailed threat intelligence report yet documents eight months of disrupted abuse: Chinese labs distilling billions of Claude exchanges, AI-orchestrated Russian espionage, an anti-torpedo weapons specification, and five potential bioweapon-use cases.
-
Policy 中150 頁的威脅清單：Anthropic 九月報告點名 Zhipu、小米、ShinyHunters 與一份解放軍海軍武器提案
Anthropic 迄今最詳盡的威脅情資報告，記錄八個月內遭到瓦解的濫用行為：中國實驗室蒸餾數十億次 Claude 對話、AI 編排的俄羅斯間諜行動、反魚雷武器規格書，以及五起潛在生化武器用途案例。
-
Models EN890 Bytes Per Token: How DeepSeek-V4.1-Flash Crushed KV Cache Costs and Retired Its Own Flagship
DeepSeek's new 552B-parameter open-weight model uses a Causal Encoder-Decoder architecture to cut KV cache to 890 bytes per token — 1/4 of the previous generation — while beating V4-Pro on agentic benchmarks, prompting DeepSeek to phase out its own flagship.
-
Models 中每 token 僅 890 位元組：DeepSeek-V4.1-Flash 如何把 KV 快取成本壓到極限，還讓自家旗艦提前退役
DeepSeek 新推出的 552B 參數開放權重模型採用因果編碼器—解碼器架構，將 KV 快取壓到每 token 890 位元組（僅前代的 1/4），同時在 Agent 基準上超越 V4-Pro，促使 DeepSeek 親手淘汰自家旗艦。
-
Tools ENA Neck, Cooler Motors and a $14,000 Tag: Unitree's G1+ Makes Humanoids Practical
Unitree's G1+ adds a 2-DOF neck, 110% more shoulder torque, 72% cooler motors and far-field voice to its $14,000 humanoid — the platform refines itself toward real work.
-
Tools EN會轉頭、更耐熱、只要 1.4 萬美元：宇樹 G1+ 讓人形機器人走向實用
宇樹科技發表 G1+ 人形機器人：新增 2 自由度頸部、肩部扭矩提升 110%、發熱降低 72%，加上遠場語音互動，標準版售價 9.5 萬人民幣（約 1.4 萬美元）。
-
Industry ENFive Years of Power Bills Up Front? Samsung and SK Hynix Just Told KEPCO No
Korea's debt-laden state utility asked its two biggest chipmakers to prepay 25 trillion won ($18.4B) of electricity bills to fund grid upgrades for the Yongin and Honam chip clusters. Both said no — leaving KEPCO's 210 trillion won debt mountain and Korea's AI power crunch unresolved.
-
Industry 中預繳五年電費？三星與 SK 海力士正式拒絕 KEPCO 的 25 兆韓元提案
負債累累的韓國電力公社要求兩大晶片廠預繳 25 兆韓元（約 184 億美元）電費，為龍仁與湖南晶片聚落的電網升級籌資。兩家都說不——KEPCO 的 210 兆韓元債務大山與韓國的 AI 電力缺口，依然無解。
-
Policy EN'Fearmongering' Meets 'Cold War Playbook': Beijing Formally Rejects the US AI Slowdown
China's Foreign Ministry and state media dismiss Amodei's pacing plan as containment dressed as safety, days before mid-September US–China AI talks — the first formal state-level reply to Silicon Valley's slowdown push.
-
Policy 中「恐嚇行銷」對上「冷戰劇本」：北京正式拒絕美國的 AI 減速倡議
中國外交部與官媒將 Amodei 的減速計畫定性為「包著安全外衣的圍堵」，發表於美中 AI 會談前夕——這是北京對矽谷減速浪潮首次正式的國家層級回應。
-
Policy ENTen Days From Ask to Workstream: China's Data Regulator Moves to Write Embodied-AI Data Standards
China's National Data Administration will develop standards for embodied-AI training data, ten days after seven robotics firms asked for them — aiming at a 10-million-hour data gap.
-
-
Industry ENNegative-Yield Convexity: Why Hong Kong Just Handed Z.AI $5 Billion for Free
Z.AI priced $3B of zero-coupon convertible bonds at a negative yield and placed $2B of stock in one move — the most aggressive capital raise in the LLM industry's short history.
-
Industry 中負殖利率的豪賭：香港為何免費奉上 50 億美元給 Z.AI
Z.AI 一口氣完成 20 億美元配股加上 30 億美元零息可轉債——定價隱含負殖利率——這是 LLM 產業短短歷史中最激進的一次募資。
-
Industry EN172 Million Shares at HK$32.96: Inside Ligent's $727M Hong Kong IPO and the Optical Land Grab Fueling AI Data Centers
Chinese optical-module maker Ligent Technologies prices its Hong Kong IPO at HK$5.7B (~$727M) with a Sept 22 debut, betting that GPU-to-GPU interconnect demand — 70% of its revenue — keeps compounding as AI clusters scale.
-
Industry EN1.72 億股、每股 32.96 港元：Ligent 7.27 億美元香港 IPO 與 AI 資料中心背後的光學軍備競賽
中國光學模組廠 Ligent Technologies 以 57 億港元（約 7.27 億美元）在香港 IPO 定價，預計 9 月 22 日掛牌，押注 GPU 互連需求持續爆發——這塊業務占其營收七成。
-
Policy ENNaming the Enemy's Models: China's Spy Chief Declares AI a 'New Arena for Strategic Rivalry'
China's State Security Minister Chen Yixin names Claude Mythos and GPT-5.5-Cyber as cyber-threats in a party journal — the first time Western frontier models appear by brand in Beijing's national-security discourse.
-
Policy 中點名敵手的模型：中國情報首長宣告 AI 是「大國戰略競爭的新競技場」
中國國家安全部部長陳一新在黨刊點名 Claude Mythos 與 GPT-5.5-Cyber 為資安威脅——西方前沿模型首次以品牌名稱出現在北京的國安論述中。
-
Industry ENSelling First, Asking Later: Asia's AI Stocks Crash as the Industry's Own CEOs Call for a Slowdown
SoftBank fell 13.2% and Kioxia 9.8% at Monday's open in Tokyo after Amodei, Altman and Musk backed a frontier development pause — the first time AI's biggest bulls have moved the market against themselves.
-
Industry 中先賣再說：AI 巨頭自己喊煞車，亞洲 AI 概念股應聲重挫
Amodei、Altman 與 Musk 表態支持放緩前沿模型開發後，SoftBank 週一開盤暴跌 13.2%、Kioxia 重挫 9.8%——這是 AI 多頭首度親手把自己的行情賣下去。
-
Models ENVision as a Feedback Loop, Not Just an Input: Ant Group Open-Sources the 124B Ling-3.0-flash-VL Under MIT
Ant Group's inclusionAI has open-sourced Ling-3.0-flash-VL, a 124B-parameter multimodal MoE that activates only 5.5B per token, reads images and video over a 256K context, ships in BF16 and FP8, and carries a permissive MIT license.
-
Models 中視覺是回饋迴路，而不只是輸入：螞蟻集團以 MIT 授權開源 124B 的 Ling-3.0-flash-VL
螞蟻集團旗下 inclusionAI 開源 Ling-3.0-flash-VL：124B 參數的多模態 MoE 模型，每 token 僅啟動 5.5B 參數，支援影像與影片輸入、256K 上下文，提供 BF16 與 FP8 權重，並採用寬鬆的 MIT 授權。
-
Industry ENFive Initiatives, One Open-Source Pitch: Xi Takes China's AI Vision to BRICS in Delhi
At the 18th BRICS Summit in New Delhi, Xi Jinping proposed a China-led BRICS AI open-source community, pitched membership in Beijing's World AI Cooperation Organization, and made open weights the diplomatic currency of the Global South.
-
Industry 中五項倡議、一場開源提案：習近平把中國的 AI 願景帶進德里 BRICS 峰會
在第 18 屆 BRICS 新德里峰會上，習近平提出由中國主導的 BRICS AI 開源社群，邀請各國加入北京的世界人工智慧合作組織，讓開放權重成為全球南方的外交貨幣。
-
Industry ENFolding Silicon Instead of Shrinking It: Huawei's Kirin 9050 Pro Puts LogicFolding on Sale
Huawei's Kirin 9050 Pro is the first mass-market chip built on LogicFolding — stacking logic in vertical tiers to lift transistor density 55% without EUV lithography. It shipped inside the Mate XT 2 tri-fold this weekend.
-
Industry 中不微縮、改折疊：華為 Kirin 9050 Pro 讓 LogicFolding 晶片首次量產上街
華為 Kirin 9050 Pro 是首款採用 LogicFolding 架構的量產晶片——透過垂直堆疊邏輯層，在完全不依賴 EUV 微影技術下將電晶體密度提升 55%，本週末隨 Mate XT 2 三折機正式開賣。
-
Models ENThe Sunset That Wasn't: DeepSeek V4.1-Flash Ships With 1M Context and $0.003 Cache Reads — and V4 Pro Gets a Reprieve
DeepSeek's new 552B-parameter open-weight model undercuts rivals on inference pricing just as the company walks back its plan to retire the V4 Pro API on September 14.
-
Models 中一場沒有發生的日落：DeepSeek V4.1-Flash 以百萬級上下文與 0.003 美元快取讀取登場，V4 Pro 獲得緩刑
DeepSeek 這款 552B 參數的開放權重新模型以極低的推理價格挑戰對手，同時公司收回於 9 月 14 日關閉 V4 Pro API 的計畫。
-
Industry ENSqueezing Tokens: How Chinese AI Labs Are Closing the Gap by Being More Efficient
A new Fortune analysis argues China's labs are matching US frontier AI not by outspending it but by out-engineering it — cheaper attention algorithms, open weights, and enterprise adoption data that show the gap has narrowed to 2.7%.
-
Industry 中把 Token 榨到極致：中國 AI 實驗室靠「效率」追平美國的真相
Fortune 九月十三日分析指出，中國實驗室並非只靠蒸餾美國模型——在算力封鎖下被迫練出的效率工程，才是它們把差距縮到 2.7%、並開始反過來拿下美國企業客戶的真正武器。
-
Meta ENNo Human in the Loop: Anthropic Says Claude Was Used to Build Autonomous Kamikaze Drone Swarms
Anthropic's September threat report reveals Russia-linked actors built FPV kamikaze drone swarms with on-board AI target selection, Houthi operators used Claude as a missile software engineer, and a Chinese EW suite defaulted to Taiwan scenarios — commercial AI has entered the weapons chain.
-
Meta 中無人批准即可開火：Anthropic 披露 Claude 曾被用來打造自主神風無人機蜂群
Anthropic 九月威脅報告揭露：俄羅斯相關行為者以 Claude 打造可自主選擇人員目標並引爆炸藥的 FPV 神風無人機蜂群、葉門青年運動操作者把 Claude 當成飛彈軟體工程師、中國電戰模組預設情境指向台灣——商用 AI 已進入武器工程鏈。
-
Policy ENWithin Legal Requirements: US Defends Pax Silica Hub as Philippine Protests Reach the Corridor's Doorstep
The State Department says the 4,000-acre Pax Silica AI and semiconductor hub in New Clark City will operate within legal requirements, as protests, relocation disputes, and Senate scrutiny put the flagship initiative on the defensive.
-
Policy 中「符合法律規範」：美國為 Pax Silica 計畫背書，菲律賓抗議群眾已走上走廊大門
美國國務院表示，規劃於新克拉克市的 4,000 英畝 Pax Silica 半導體與 AI 中心將「符合法律規範」運作。然而抗議行動、徵地補償爭議與參議院調查，正讓這項旗艦計畫承受空前壓力。
-
Policy ENThirty Years for a Stolen Blueprint: South Korea's Rewritten Espionage Law Takes Effect Tomorrow
On September 13, South Korea's first espionage-law rewrite in 73 years takes effect — extending spying charges beyond North Korea to any foreign beneficiary, with courts empowered to impose up to 30 years for leaking chip technology. The target is China's recruitment of Samsung and SK Hynix engineers.
-
Policy 中竊取一張藍圖，坐牢三十年：南韓 73 年來首次改寫的間諜法明天生效
9 月 13 日，南韓 73 年來首次大幅改寫的刑法第 98 條之 2 正式生效——間諜罪不再限於北韓受益，擴及任何外國或同等組織，法院最高可判 30 年。目標直指中國對三星與 SK 海力士工程師的挖角行動。
-
Policy EN'No, I Don't Have Any': Trump Dismisses AI Extinction Risk as Geneva Talks Loom
Days before the first dedicated US-China AI safety dialogue, President Trump brushed aside extinction-risk warnings and framed China winning the AI race as the only danger that matters.
-
Policy 中「不，我沒有任何擔憂」：日內瓦會談前夕，川普駁回 AI 滅絕風險論
在美中首場專責 AI 安全對話登場前幾天，川普總統公開表示不擔心 AI 滅絕人類的風險，並把「輸給中國」定義為唯一值得在意的人工智慧威脅。
-
Meta ENAgents Gone Wild: How Hundreds of AI Agents Hacked 395 Organizations in 48 Countries
A likely Russian-speaking attacker used hundreds of autonomous AI agents to exploit PaperCut flaws, breaching 440 servers across 48 countries — and 11 orgs fell in 26 seconds. The agents even ignored their operator's targeting rules.
-
Meta 中代理人失控：數百個 AI 代理如何攻陷 48 國 395 個組織
一名疑似俄語攻擊者動用數百個自主 AI 代理攻擊 PaperCut 漏洞，在全球 48 國攻陷 440 台伺服器——26 秒內就有 11 個組織淪陷，代理甚至無視操作者的下手指令。
-
Models EN8B Active Parameters and a 890-Byte Cache: DeepSeek's V4.1-Flash Outscores Its Own Flagship — and Phases It Out
DeepSeek's new 552B-parameter MoE activates just 8B parameters per input token, compresses its KV cache to 890 bytes per token, beats V4-Pro on Terminal-Bench 2.1 and DeepSWE — and inherits the flagship's API traffic on September 14.
-
Models EN每 token 僅啟動 8B 參數、KV 快取壓到 890 位元組：DeepSeek V4.1-Flash 全面超越自家旗艦——然後把它淘汰
DeepSeek 新發布的 552B 參數 MoE 模型，每個輸入 token 僅啟動 8B 參數，KV 快取壓縮至每 token 890 位元組，在 Terminal-Bench 2.1 與 DeepSWE 上擊敗 V4-Pro——並將於 9 月 14 日承接旗艦模型的 API 流量。
-
Meta ENYou Thought You Were Talking to Kimi: Inside the Relay Scheme That Served Claude to Millions
Anthropic's threat report reveals Moonshot and DeepSeek silently forwarded real customer requests to Claude and showed the answers as their own — 151 million exchanges for Alibaba, cross-session replay attacks against encrypted reasoning, and developer secrets from a dozen countries caught in the middle.
-
Meta 中你以為在跟 Kimi 對話：起底把 Claude 偷偷送進數百萬用戶對話的中轉計畫
Anthropic 威脅報告揭露，Moonshot 與 DeepSeek 曾把真實客戶的請求悄悄轉發給 Claude，再把答案當成自家模型的回覆——Alibaba 量產 1.51 億次對話、跨 session 重放攻擊破解加密推理、十幾國開發者的商業機密全被捲入。
-
Industry ENFrom $300M to $1B in Two Months: Moonshot's Kimi K3 Just Rewrote the Open-Weight Business Model
Moonshot AI told investors its annualized revenue crossed $1B in August — up from $300M in June — driven by Kimi K3, the largest open-weight model ever released. It now targets $2B ARR by year-end ahead of a Hong Kong IPO.
-
Industry 中兩個月從 3 億到 10 億美元：Moonshot 的 Kimi K3 改寫了開放權重商業模式
Moonshot AI 向投資人透露，其年化營收在 8 月突破 10 億美元——較 6 月的 3 億美元大幅躍升，幕後推手是史上最大開放權重模型 Kimi K3。在香港 IPO 之前，公司目標年底達到 20 億美元 ARR。
-
Industry ENFrom 142 to 410 Yuan in One Bell: Enflame's Debut Triples the Last AI Dragon's valuation
Enflame opened 188% above its IPO price on the Shanghai STAR Market, instantly valuing Tencent's captive AI-chip maker at 176.4 billion yuan — and completing the public-market coronation of China's four AI-chip dragons.
-
Industry 中一聲鐘響從 142 到 410 元：燧原科技掛牌暴漲，最後一條 AI 晶片小龍完成三倍市值加冕
燧原科技在上海科創板掛牌首日以 410 元開盤、較發行價大漲 188%，騰訊加持的這家 AI 晶片商市值一舉衝上 1,764 億人民幣 — 中國四大 AI 晶片「小龍」的資本市場加冕至此全數完成。
-
Models EN8B Open Weights Beat the Closed Frontier: SenseTime's SenseNova-U1.5 Outscores Nano-Banana-Pro on Vision Reasoning
SenseTime's 8B-parameter SenseNova-U1.5-8B-MoT natively unifies visual understanding and generation without encoders or VAEs — and posts 68.2% on VBVR-Pro-Bench, ahead of proprietary Nano-Banana-Pro at 56.4%, under Apache 2.0.
-
Models 中80 億參數開源擊敗閉源前沿：商湯 SenseNova-U1.5 視覺推理評測超越 Nano-Banana-Pro
商湯以 Apache 2.0 釋出的 8B 原生統一多模態模型 SenseNova-U1.5-8B-MoT，在 VBVR-Pro-Bench 視覺推理評測拿下 68.2%，超越專有模型 Nano-Banana-Pro 的 56.4%，並原生支援 4K 生成。
-
Models EN64% on Terminal-Bench From a 122B MoE: How T1 Turned Agent RL Into an Infrastructure Problem
Tencent Hy's T1 post-trains Qwen3.5-122B-A10B with pure verifier-driven RL to reach 64.0% on Terminal-Bench 2.1 — and its real contribution is the unglamorous plumbing: TITO token-faithful training, R3 routing replay, and a dense assertion-count reward.
-
Models 中122B MoE 終端 Agent 拿下 Terminal-Bench 64%：T1 把 Agent 強化學習變成一場基礎設施工程
騰訊 Hy 團隊的 T1 以 Qwen3.5-122B-A10B 為底座、純用驗證器驅動的強化學習後訓練，在 Terminal-Bench 2.1 拿下 64.0%——而它真正的貢獻是那些不性感的管線工程：TITO token 保真訓練、R³ 路由回放，與按斷言計數的稠密獎勵。
-
Meta ENSpies, Scammers and Bioweapon Labs: Inside Anthropic's Most Disturbing Threat Report Yet
Anthropic's September 2026 threat intelligence report documents Russian AI-driven espionage with self-rebuilding malware, Chinese dissident-surveillance pipelines, and five bioweapon-adjacent research cases — the first such disclosure by any AI company.
-
Meta 中間諜、詐騙集團與生物武器實驗室：解析 Anthropic 迄今最駭人的威脅報告
Anthropic 2026 年 9 月威脅情資報告揭露俄羅斯 AI 自動化間諜行動（惡意軟體會自我改寫躲避偵測）、中國異議人士監控管線，以及五起涉及生物武器的 research 案例——這是 AI 公司首度公開此類證據。
-
Models EN124B Parameters, Open Weights: Ant Group Open-Sources Ling-3.0-flash-Fin and the FinFIRST Benchmark
Ant Group releases Ling-3.0-flash-Fin, a finance-tuned MoE model with 124B total / 5.1B active parameters plus the expert-built FinFIRST benchmark — betting that Wall Street's next research analyst runs on open weights with auditable evidence chains.
-
Models 中1,240 億參數、開放權重：螞蟻集團開源 Ling-3.0-flash-Fin 與 FinFIRST 金融代理人基準
螞蟻集團開源金融強化模型 Ling-3.0-flash-Fin（124B 總參數／每 token 僅啟動 5.1B 的 MoE 架構），連同與中金公司投行團隊共同打造的專家級基準 FinFIRST——押注華爾街下一代研究分析師，將運行在可稽核證據鏈的開放權重模型上。
-
Models ENThe Flash That Ate the Flagship: DeepSeek V4.1 Flash Arrives — and Retires V4 Pro
DeepSeek's new 552B MoE with native multimodality beats its own flagship on agent benchmarks — so the company is routing all V4 Pro traffic to it, at Flash prices, from September 14.
-
Models 中吞噬旗艦的 Flash：DeepSeek V4.1 Flash 正式登場——並讓 V4 Pro 退役
DeepSeek 這款 552B 參數、原生多模態的新 MoE 模型在 agent 基準測試上擊敗自家旗艦——公司因此宣布 9 月 14 日起所有 V4 Pro 流量改由它承接，並以 Flash 價格計費。
-
Research EN16.79% to 34.97%: The Prefill Experiment That Just Put Qwen 3.8's Training Data on Trial
An independent reasoning-prefill analysis found Qwen 3.8's answer overlap with GPT-5.5 Pro more than doubles when seeded with 1% of GPT's chain-of-thought — the sharpest technical signal yet in the distillation debate.
-
Research 中從 16.79% 到 34.97%：一場預填充實驗，讓 Qwen 3.8 的訓練資料站上被告席
獨立研究者的推理預填充分析發現：注入 GPT-5.5 Pro 前百分之一的思考鏈後，Qwen 3.8 的答案重疊率翻倍——這是蒸餾爭議至今最尖銳的技術訊號。
-
Industry EN3 Million Robots, 1 Million Autonomous Vehicles, 100,000 Drones: JD.com Goes All-In on Physical AI
At JDDiscovery 2026 in Beijing, JD.com launched a Physical AI Acceleration Plan targeting six 'world's largest' milestones, anchored by JD Logistics' plan to procure 3 million robots over five years.
-
Industry 中300 萬機器人、100 萬自駕車、10 萬無人機：京東全面押注實體 AI
京東於北京 JDDiscovery 2026 發布「實體 AI 加速計畫」，鎖定六個「全球最大」目標，核心是京東物流五年內採購 300 萬台機器人的藍圖。
-
Industry EN20-50% Price Hikes Across the Board: HBM Shortage Squeezes China's AI Chipmakers
Reuters reports Huawei, Cambricon, MetaX and Iluvatar CoreX raised AI accelerator prices 20-50% in two months as the HBM shortage bites — Huawei's Ascend 950DT now quotes above 250,000 yuan.
-
Industry 中全系漲價 20-50%：HBM 短缺重擊中國 AI 晶片商
路透社報導，華為、寒武紀、摩爾線程與壁仞等中國 AI 晶片商在兩個月內將加速卡價格調漲 20-50%，華為 Ascend 950DT 報價已突破 25 萬人民幣。
-
Industry ENCheaper Than Open Source: OpenAI Pitches AI for Chip Design and Claims It Undercuts China on Price
At Goldman Sachs' Communacopia conference, CFO Sarah Friar said OpenAI's 80% Luna price cut drove 10x usage — and that its API can now beat Chinese open-source models on total cost in specialized verticals like chip design and life sciences.
-
Industry 中比開源還便宜？OpenAI 把 AI 推進晶片設計，宣稱在價格上擊敗中國對手
在高盛 Communacopia 科技會議上，OpenAI 財務長 Sarah Friar 表示 Luna 降價 80% 帶來十倍用量成長，並宣稱在晶片設計、生命科學等專業領域，其 API 的總持有成本已低於中國開源模型。
-
Meta EN100,000 Homegrown GPUs: JD Cloud and Moore Threads to Build China's Largest Domestic-Chip AI Cluster
At its Global Technology Explorers Conference, JD Cloud committed to a 100,000-GPU intelligent-computing cluster built entirely on Moore Threads' domestic full-function GPUs — the first time Chinese silicon reaches 100K scale inside a top AI cloud provider.
-
Meta EN十萬張國產 GPU：京東雲攜手摩爾線程，打造中國最大全國產晶片 AI 叢集
在 2026 京東全球科技探索者大會上，京東雲宣布擬建設十萬卡全功能 GPU 智算叢集，算力底座全面採用摩爾線程國產 GPU——這是國產晶片首次進入頭部 AI 雲廠商的十萬卡級核心叢集。
-
Policy ENSelf-Reliance, Not Stolen Tokens: Beijing Formally Rejects the US Distillation Advisory
China's Foreign Ministry calls the NSA-CISA-FBI accusations 'false allegations' and hails its AI as homegrown, escalating a war of words days before Trump-Xi talks where AI governance tops the agenda.
-
Policy 中自主研發，而非竊取：北京正式駁斥美國模型蒸餾指控
中國外交部稱 NSA、CISA 與 FBI 的聯合指控為「不實指控」，強調中國 AI 成就源於自主研發。這場交鋒距離川習會僅剩數週，AI 治理預計是核心議題。
-
Industry ENBeijing Pulls Up the IPO Drawbridge: CSRC 'Window Guidance' Hits Humanoid-Robot Listings After Unitree's Crash
China's securities regulator is informally telling banks and startups that humanoid-robot IPOs now require proven recurring revenue and a path to profitability — a direct response to Unitree's 5x debut pop and 45% collapse.
-
Industry 中北京拉起 IPO 吊橋：宇樹崩跌後，證監會「窗口指導」收緊人形機器人上市
中國證監會以非正式的「窗口指導」通知投行與企業：人形機器人 IPO 如今必須證明有經常性收入、虧損持續收斂——直接回應宇樹上市首日暴漲逾 5 倍又崩跌 45% 的教訓。
-
Meta ENSix Hours, Thousands of Credentials: Google's GTIG Documents the Shift From Prompting to Agentic Attacks
Google's latest AI Threat Tracker chronicles an autonomous multi-agent credential harvest that compromised thousands of logins in under six hours — and a broad pivot by adversaries toward AI assets and agent-driven operations.
-
Meta 中六小時、數千組憑證：Google GTIG 揭露攻擊者從「下提示」走向「代理化作戰」
Google 最新 AI 威脅追蹤報告記錄了一場自主多代理憑證竊取行動——在六小時內攻陷數千組帳密，並揭露攻擊者全面轉向 AI 資產與代理化攻擊的趨勢。
-
Models ENA Flash Tier Retiring the Pro: Inside DeepSeek's V4.1 Flash Beta and Its 48-Hour Deadline
DeepSeek quietly opened a two-day beta for V4.1 Flash — a new-architecture, natively multimodal model that claims to surpass V4 Pro at Flash pricing, with outputs peaking at 507 tokens/s.
-
Models 中用 Flash 退役 Pro：DeepSeek V4.1 Flash 限時 48 小時公開測試全解析
DeepSeek 低調開放 V4.1 Flash 兩天限時測試——全新架構、原生多模態，宣稱以 Flash 價格超越 V4 Pro，輸出速度實測最高達每秒 507 tokens。
-
Research EN49 Billion Compounds, Nine Finalists, One Approval: Inside Mprosevir, China's First AI-Assisted Class 1 Drug
China's NMPA has granted conditional approval to Mprosevir, an AI-assisted COVID-19 antiviral that went from candidate discovery to completed clinical trials in 3.5 years — the world's first approved small molecule born from DNA-encoded library technology.
-
Research EN49 億化合物、9 個入圍者、1 張藥證：解析中國首款 AI 輔助一級新藥 Mprosevir
中國國家藥監局有條件核准 Mprosevir——一款 AI 輔助開發的 COVID-19 抗病毒藥物，從候選藥物發現到臨床試驗完成僅花 3.5 年，也是全球首款源自 DNA 編碼文庫技術獲批的小分子新藥。
-
-
Meta 中從舞台到戰場：中國解放軍加速將人形機器人推向實戰
-
Models ENFrom Simulating the World to Acting in It: HiDream's O1-Embodied Tops RoboColiseum's Hardest Leaderboard
HiDream.ai launches HiDream-O1-Embodied, an embodied world model that ranks No. 1 on RoboColiseum's Robustness leaderboard (0.692) and pairs a 100x generative data flywheel with Noitom motion capture.
-
Models 中從模擬世界到親手操作：HiDream 推出 O1-Embodied，奪下 RoboColiseum 最難排行榜榜首
HiDream.ai 發布具身世界模型 HiDream-O1-Embodied，以 0.692 分登上 RoboColiseum 魯棒性排行榜第一名，並以 Noitom 動捕資料結合 100 倍生成式擴增打造資料飛輪。
-
Policy ENSix Labs, Billions of Tokens: NSA, CISA and FBI Formally Accuse China's AI Firms of Industrial-Scale Model Distillation
Joint advisory AA26-251A names DeepSeek, Moonshot AI, Alibaba, MiniMax, StepFun and Z.AI, alleging they distilled billions of tokens from Claude, GPT, Gemini and Grok since late 2024 — and prescribes a quiet-poison playbook for defenders.
-
Policy 中六家實驗室、數十億 token：NSA、CISA 與 FBI 正式指控中國 AI 公司大規模蒸餾美國模型
聯合公告 AA26-251A 點名 DeepSeek、Moonshot AI、阿里巴巴、MiniMax、StepFun 與 Z.AI，指控他們自 2024 年底以來從 Claude、GPT、Gemini 與 Grok 蒸餾數十億 token，並為防禦方開出了一套「靜默投毒」劇本。
-
Policy ENSix Chinese AI Firms Named and Shamed: Inside the NSA-CISA-FBI Advisory on Industrial-Scale Model Distillation
Joint advisory AA26-251A alleges DeepSeek, Moonshot AI, Alibaba, MiniMax, StepFun and Z.AI extracted billions of tokens from Claude, GPT, Gemini and Grok since late 2024 — and recommends quietly serving degraded responses to suspected distillation traffic.
-
Policy EN9,800 EFLOPS by 2030: China's ¥3.8 Trillion Compute Plan Turns Ambition Into a Procurement Curve
MIIT's 15th Five-Year Plan for ICT quadruples China's intelligent computing target to 9,800 EFLOPS by 2030, backed by ¥3.8 trillion in cumulative information infrastructure investment and 100,000-card clusters.
-
Policy 中2030 年 9,800 EFLOPS：中國 3.8 兆元算力藍圖，把雄心變成一條採購曲線
工信部「十五五」資通訊規劃將中國智慧算力目標定為 2030 年 9,800 EFLOPS（現況四倍以上），配套 3.8 兆人民幣資訊基礎設施累計投資與十萬卡叢集部署。
-
Policy ENFirst Mover on Huawei AI Silicon: Malaysia Leans Toward Ascend 910C for Its RM2 Billion Sovereign AI Push
Malaysia is seriously evaluating Huawei Ascend 910C chips as the backbone of a RM2 billion sovereign AI program, defying explicit US warnings that using the chips could breach export regulations.
-
Policy 中華為 AI 晶片的首位國家級買家？馬來西亞傾向以 Ascend 910C 打造 20 億令吉主權 AI 基礎設施
馬來西亞正認真評估以華為 Ascend 910C 晶片作為 20 億令吉主權 AI 計畫的核心，無視美方明確警告使用該晶片可能違反美國出口管制。
-
Policy EN9,800 EFLOPS by 2030: China's Five-Year Plan Commits $532 Billion to Quadrupling AI Compute
China's MIIT has set a target of 9,800 EFLOPS of intelligent computing capacity by 2030 — more than four times today's level — backed by 3.8 trillion yuan in planned information infrastructure investment.
-
Policy 中2030 年 9,800 EFLOPS：中國五年計劃投入 5,320 億美元，要將 AI 算力翻四倍
中國工信部發布「十五五」資訊通信業發展規劃，設定 2030 年智能算力規模達 9,800 EFLOPS 的目標——超過目前水準的四倍——並規劃 3.8 兆元人民幣的資訊基礎設施累計投資。
-
Meta ENHundreds of Millions of Accounts, Zero Clicks: AI Models Built the WeWorm Attack on WeChat
California researchers used AI models to construct WeWorm, a self-propagating zero-click worm that could have hijacked hundreds of millions of WeChat accounts within hours — reading messages, sending texts, and making calls as the victim. Tencent says it has already patched the flaw.
-
Meta 中數億帳號、零點擊：AI 模型打造出針對微信的 WeWorm 攻擊
加州研究人員利用 AI 模型建構出 WeWorm——一種零點擊、可自我傳播的電腦蠕蟲，能在數小時內劫持數億個微信帳號，以受害者身分讀取訊息、發送訊息甚至撥打電話。騰訊表示漏洞已完成修補。
-
Models ENNo Teleop, No Script: Unitree's UnifoLM-X2-1.0 Drives Fully Autonomous Humanoid Combat With a Real-Time World Model
Unitree claims UnifoLM-X2-1.0 is the first real-time world-model system to control fully autonomous humanoid combat — no teleoperation, no choreography. Inside the claim, the WMA-0 lineage, and what's still unproven.
-
Models 中無遙控、無腳本：宇樹 UnifoLM-X2-1.0 用即時世界模型驅動全自主人形機器人格鬥
宇樹科技宣稱 UnifoLM-X2-1.0 是首個以即時世界模型驅動全自主人形機器人格鬥的系統——沒有遙控、沒有預編排動作。深入解析這項宣告、WMA-0 的技術血統，以及尚待驗證的疑點。
-
Industry ENEurope's Biggest Round Ever: Mistral Raises €3B Led by Samsung to Build the Sovereign AI Stack
Mistral AI's €3B Series D at a €21B+ post-money valuation is the largest equity round in European tech history — Samsung leads, EQT's Scaleup Europe Fund and PSG Equity co-lead, and the bet is on sovereign, open-weight AI.
-
Industry 中歐洲史上最大輪融資：Mistral 獲三星領投 30 億歐元，打造主權 AI 全堆疊
Mistral AI 以超過 210 億歐元估值完成 30 億歐元 Series D，創歐洲科技公司股權融資紀錄——三星領投、EQT 與 PSG 共同領投，押注主權開源權重 AI。
-
Tools ENFrom Voice Commands to a Household Agent: Baidu's Xiaodu Takes Its 'Family AI Brain' to Hardware
At its September 8 launch event in Beijing, Baidu's Xiaodu unveiled an agent-first hardware lineup — Tiantian companion screens, smart displays, speakers, and cameras running the agentic 'Chaoneng Xiaodu' assistant and a 2.0 AI caregiving agent.
-
Tools 中從語音指令到家庭智能體：百度小度把「家庭 AI 大腦」裝進硬體
百度小度於 9 月 8 日北京發表會推出以智能體為核心的硬體陣容——添添閨蜜機、智慧螢幕、音箱與攝影機，全面搭載智能體化的「超能小度」，攝影機並迎來第二代 AI 看護智能體。
-
Industry ENFour Logos, One Framework: Alibaba Cloud, Cambricon, and Ant Group Take Seats in the PyTorch Foundation
At PyTorch Conference China in Shanghai, the PyTorch Foundation welcomed Alibaba Cloud and Cambricon as Platinum members and Ant Group as Gold — putting Chinese cloud, chip, and fintech giants on its governing board alongside Huawei, as more than 250 Chinese organizations now contribute to PyTorch ecosystem projects.
-
Industry 中四家巨頭、一個框架：阿里雲、寒武紀與螞蟻集團正式入座 PyTorch 基金會
在上海舉行的 PyTorch Conference China 上，PyTorch 基金會宣布阿里雲與寒武紀成為白金會員、螞蟻集團成為黃金會員，與華為同台——超過 250 家中國機構如今參與 PyTorch 生態系專案，開源 AI 基礎設施的治理版圖正在重畫。
-
Industry ENThe Broker of Last Resort: Huawei Now Bankrolls China's Lithography Startups and Matchmakes Their SMIC Deals
A Financial Times investigation reveals Huawei is funding Chinese lithography equipment makers through its Hubble arm and personally brokering their sales to SMIC — the clearest sign yet that the sanctioned giant has decided if it cannot buy the machines, it will own the industry that makes them.
-
Industry 中最後掮客現身：華為出資中國微影設備新創，還親自撮合它們與中芯的交易
英國《金融時報》調查揭露，華為正透過哈勃投資注資中國微影曝光設備廠商，並親自居中撮合這些廠商與中芯國際的採購協議——這是這家被制裁巨頭至今最明確的訊號：買不到的機器，就乾脆買下整個製造這些機器的產業。
-
Models ENA 2B Model That Beats the 4B Class: OpenBMB Open-Sources MiniCPM5-2B With 131K Context and a Fully Open Data Stack
OpenBMB's MiniCPM5-2B averages 53.9 across 34 benchmarks, outscoring every 4B-class rival in its comparison set, and ships with 131K context, hybrid thinking, and the entire UltraData training stack under Apache 2.0.
-
Models 中2B 模型打贏 4B 級對手：OpenBMB 開源 MiniCPM5-2B，131K 上下文 plus 全套資料集一次奉送
OpenBMB 的 MiniCPM5-2B 在 34 項基準測試平均拿下 53.9 分，超越對比組中所有 4B 級模型，並以 Apache 2.0 授權釋出 131K 上下文、混合思考模式與完整 UltraData 訓練資料棧。
-
Industry ENOne Flight to Beijing, Four Hundred Lost Jobs: Belgium Charges a Former BelGaN Executive With GaN Chip Espionage
Belgian federal prosecutors say a 52-year-old dual national spent three years as a senior BelGaN researcher while directing a Chinese look-alike startup — the clearest public blueprint yet of how gallium-nitride know-how left Europe's last GaN fab.
-
Industry 中一趟北京航班與四百個失去的工作：比利時以間諜罪起訴 BelGaN 前高層，氮化鎵技術外洩案曝光
比利時聯邦檢方指出，一名 52 歲雙重國籍男子一邊在 BelGaN 擔任研究高層，一邊遙控中國的「鏡像公司」——這是歐洲最後一座氮化鎵晶圓廠技術外流至今最清晰的路徑圖。
-
Policy ENEight Years Later, Huawei Finally Faces a Brooklyn Jury: Inside the RICO Trial That Could Redraw US–China Tech Lines
Jury selection begins September 8 in the long-delayed US racketeering trial of Huawei — a 16-count case spanning Iran sanctions evasion, HSBC bank fraud, and trade-secret theft from six US competitors.
-
Policy 中八年纏訟終於開庭：華為 RICO 刑事審判在布魯克林展開，美中科技戰的司法攤牌時刻
美國政府對華為的刑事詐欺與智慧財產權竊盜案於 9 月 8 日在紐約布魯克林聯邦法院展開陪審團遴選，16 項罪名涵蓋伊朗制裁規避、匯豐銀行詐欺與六家美國競爭對手的營業秘密竊取。
-
Models ENOne Brain for the Whole Car: Alibaba Open-Sources Qwen-Drive-1.0, a 4B VLM That Sees, Explains, and Drives
Alibaba's Qwen team has open-sourced Qwen-Drive-1.0-4B, a vision-language foundation model that unifies 3D perception, driving Q&A, and trajectory planning in one Apache 2.0 package — while keeping the base Qwen3.5-4B entirely untouched.
-
Models 中一顆大腦開整台車：阿里巴巴開源 Qwen-Drive-1.0，4B 視覺語言模型同時會看、會講、會開車
阿里巴巴 Qwen 團隊開源 Qwen-Drive-1.0-4B：一套視覺語言基礎模型，整合 3D 感知、駕駛問答與軌跡規劃於一身，採 Apache 2.0 授權釋出，且完全不改動基礎的 Qwen3.5-4B 架構。
-
Industry ENThe Founder Returns: Zhang Yiming Personally Leads ByteDance's Real-Time World Model Bid
Bloomberg reports ByteDance is building a real-time spatial-video world model on Seedance, targeting ~20 fps at ~50ms latency for cloud-rendered Pico worlds, with a launch possible as soon as next month.
-
Industry 中創辦人回歸第一線：張一鳴親自督軍，字節跳動即時世界模型劍指下月發布
彭博報導，字節跳動正基於 Seedance 打造即時空間影片世界模型，目標約 20 fps、延遲低於 50 毫秒，由雲端渲染餵養 Pico 頭顯，最快下個月推出。
-
Industry ENSix Years in Exile, One Fold Forward: Huawei's Kirin 9050 Pro Reboots the Flagship Chip Race Without US Silicon
Huawei's Mate XT2 launches with the Kirin 9050 Pro, the first flagship SoC built on its homegrown LogicFolding 3D-stacking architecture — a 42% uplift claimed with zero US supply chain.
-
Industry 中六年晶片流放後的一步：華為 Kirin 9050 Pro 用 LogicFolding 重啟旗艦晶片競賽，完全繞開美國供應鏈
Mate XT2 三折機搭載 Kirin 9050 Pro 登場 — 這是首款採用華為自研 LogicFolding 3D 堆疊架構的旗艦 SoC，官方宣稱效能提升 42%，且整機不含任何受美國管制零組件。
-
Industry EN12.7 Million Graduates, One Vanishing Entry Level: Inside China's AI Jobs Squeeze
A record 12.7 million Chinese graduates are entering a labor market where AI is absorbing the junior white-collar tasks that once trained them — and Beijing is scrambling to respond.
-
Industry 中1,270 萬畢業生，一個消失的入門職缺：透視中國的 AI 就業擠壓
中國史上最大批的 1,270 萬大學畢業生，正走進一個由 AI 接手基層白領工作的勞動市場——北京正全力尋找對策。
-
Meta EN160,000 Ascend Chips and No Nvidia in Sight: Inside DeepSeek's Inner Mongolia Mega-Cluster
Bloomberg reports DeepSeek will deploy at least 160,000 Huawei Ascend 950DT accelerators at a 1 GW data center in Inner Mongolia — the largest known all-domestic AI cluster and a stress test of China's push to replace Nvidia.
-
Meta 中16 萬顆 Ascend 晶片、零 Nvidia：拆解 DeepSeek 內蒙古超級集群
Bloomberg 報導 DeepSeek 將在內蒙古 1 GW 資料中心部署至少 16 萬顆華為 Ascend 950DT 加速器——這是已知最大規模的全國產 AI 集群，也是中國「去 Nvidia 化」路線圖最嚴苛的一次壓力測試。
-
Policy ENThe $5.6 Billion Backdoor: How Blacklisted Inspur Kept Buying Nvidia's Best Chips Through Its American Arm
An NYT investigation reveals Aivres, the California-based subsidiary of blacklisted Inspur, shipped over $3B in Blackwell servers to Southeast Asia — exposing a structural hole in US chip export controls.
-
Policy 中56 億美元的後門：被列黑的浪潮集團如何透過美國子公司持續取得 Nvidia 最強晶片
《紐約時報》調查揭露，列入實體清單的中國浪潮集團，其美國子公司 Aivres 兩年內向東南亞出口逾 30 億美元 Blackwell 伺服器，暴露美國晶片出口管制的結構性漏洞。
-
Policy ENTwo Superpowers, One Chat Window: Trump-Xi Summit Puts AI on the Sept 24 Agenda
Nikkei and Reuters reporting reveal the agenda taking shape for the first leaders-level US-China AI dialogue: monitoring AI-directed cyberattacks, voluntary lab self-policing, distillation disputes, and a Chinese bid to reopen chip export controls.
-
Policy 中兩個超級強權、一個對話視窗：川習會將 AI 推上 9 月 24 日議程
日經與路透的報導揭露了首場領導人層級美中 AI 對話的輪廓：監測 AI 網路攻擊、實驗室自律、蒸餾爭議，以及中方重啟晶片出口管制的企圖。
-
Tools EN1,500 Tokens a Second, With a Catch: What Cerebras's Qwen 3.8 27B Launch Really Tells Us About Inference Economics
Cerebras is serving Alibaba's Qwen 3.8 27B at roughly 1,500 tokens/second on wafer-scale SRAM — but a 150k TPM cap that bills cached input at full price and a 128k context ceiling make sustained agentic coding 5x pricier than slower rivals.
-
Tools 中每秒 1,500 tokens 的代價：Cerebras 上架 Qwen 3.8 27B，揭露推論經濟學的真相
Cerebras 以晶圓級 SRAM 將阿里巴巴的 Qwen 3.8 27B 推到約每秒 1,500 tokens——但快取輸入全額計費的 150k TPM 上限與 128k 上下文天花板，讓長時間代理式編碼比慢速對手貴上 5 倍。
-
Research ENThe CT Scan You Already Had: COCA Finds Colorectal Cancer Hiding in 27,000 Routine Scans
A deep learning model from Alibaba's DAMO Academy detects colorectal cancer on routine noncontrast CT with 86.6-88.2% real-world sensitivity and 99.5% specificity — and it caught 5 cancers clinicians missed.
-
Research 中你早就照過的 CT：COCA 從 2.7 萬張例行掃描中揪出大腸癌
阿里巴巴達摩院等機構研發的深度學習模型 COCA，能在例行無顯影劑 CT 上以 86.6%–88.2% 的實世界敏感度偵測大腸癌，特異度達 99.5% 以上，並抓出 5 例臨床漏診的癌症。
-
Industry ENThe Quiet Winner of IFA 2026: D-Robotics Wants to Be the Android of Home Robots
IFA's flashiest home robots — TCL's hey AiMe, Vbot's SuperDog, the xLean TR1 — all run on Sunrise AI chips from D-Robotics, the Horizon Robotics spinoff that just raised $270M to own the compute layer under embodied AI.
-
Industry 中IFA 2026 的隐形赢家：D-Robotics 想当家用机器人的「安卓」
IFA 最受瞩目的三款家用机器人——TCL hey AiMe、Vbot SuperDog、xLean TR1——全都跑在 D-Robotics 的 Sunrise AI 晶片上。这家从地平线机器人拆分出来的公司刚募得 2.7 亿美元，要拿下具身智慧的运算层。
-
Industry ENTokens With Your Dumplings: How China Turned AI Compute Into a Consumer Perk
Chinese banks, telcos, and restaurants now bundle AI tokens like credit-card points and data plans — a supply-led experiment in making compute a everyday consumer good.
-
Industry 中吃餃子送算力：中國如何把 AI Token 變成日常消費品
中國的銀行、電信業者與餐廳，正把 AI token 包裝成信用卡紅利與流量資費——一場把算力變成日常消費品的供給端實驗。
-
Industry ENSix Months Behind, Playing to Win: Kai-Fu Lee's iPhone-vs-Android Map of the US-China AI Race
In a wide-ranging Mishal Husain interview, the 01.AI founder says the US-China frontier gap has collapsed from years to six months — and that China will win AI on reach while America wins on revenue.
-
Industry 中從落后四年到只剩六個月：李開復用 iPhone vs. Android 解讀中美 AI 競賽
李開復接受 Mishal Husain 專訪時指出，中美前沿模型差距已從三、四年縮短到約六個月——中國將以觸達規模取勝，美國則賺走大部分利潤。
-
Industry ENBanned in America, Dispensing in Berlin: Galbot's G1 Robot Works Real Pharmacy Shifts at IFA
Galbot's wheeled humanoid G1 handles 5,000 SKUs in a Beijing pharmacy at 99.5% pick success and runs 170+ autonomous stores across 40 Chinese cities — but a July 28 FCC Covered List ruling bars new units from the US, turning IFA Berlin into the company's European beachhead.
-
Industry 中在美國被封殺，在柏林抓藥：Galbot G1 機器人在 IFA 現場示範真實藥局輪班
Galbot 的輪式人形機器人 G1 在北京藥局處理 5,000 種商品、撿貨成功率 99.5%，並在中國 40 多個城市營運超過 170 家無人店——但 7 月 28 日 FCC 將其硬體類別列入 Covered List，新款機型不得進入美國，IFA 柏林因此成為這家公司進軍歐洲的灘頭堡。
-
Policy ENIt Is Now or Never: Inside the First Dedicated US-China AI Safety Talks of Trump's Second Term
A Reuters exclusive reveals Washington and Beijing are finalizing mid-September talks on AI-driven cyberattacks, self-policing AI labs, and model distillation — led by Scott Bessent and timed to the Trump-Xi summit.
-
Policy 中「現在不行，就永遠沒機會了」：川普第二任期首次美中 AI 安全專門對話內幕
路透獨家披露，美中正敲定 9 月中旬的 AI 安全對話：議題涵蓋 AI 網路攻擊聯合監測、實驗室自律機制與模型蒸餾爭議，由美國財長貝森特領銜，並與川習會掛鉤。
-
Industry ENMoonshot Files for Hong Kong: Inside the Kimi Maker's $50 Billion Bid to Become China's Defining AI IPO
Moonshot AI, the Beijing startup behind Kimi, has confidentially filed for a Hong Kong IPO targeting $3–5 billion at a ~$50 billion valuation — the first major public-market test of China's AI boom, built on open weights and cloud revenue-sharing.
-
Industry 中月之暗面遞交港股上市申請：Kimi 幕後公司以 500 億美元估值，爭做中國 AI 世代定價基準
Kimi 開發商月之暗面已保密遞交香港 IPO 申請，目標募集 30 至 50 億美元、估值約 500 億美元——這是中國 AI 熱潮首次面對公開市場檢驗，靠的是開放權重策略與雲端分潤談判。
-
Industry ENSovereignty for Silicon: G42 Weighs Selling Majority Control to US Firms to Keep the Chips Coming
Bloomberg reports Abu Dhabi's G42 has held exploratory talks to sell a majority stake to American companies — betting that US ownership is the only durable way to guarantee advanced AI chip access beyond its April 2027 license-free window.
-
Industry 中以主權換晶片：G42 考慮出售多數股權給美企，只為讓輝達晶片不斷貨
彭博報導，阿布達比的 G42 已展開初步磋商，考慮將多數股權出售給美國公司——押注唯有「美資化」才能在 2027 年 4 月的免許可窗口到期後，持續取得先進 AI 晶片。
-
Industry ENFrom $10.8 Billion to $29.6 Billion in Two Years: ByteDance's Record Loan Is the AI Debt Era in One Number
ByteDance has locked in a $29.6 billion syndicated loan from nearly 30 banks to fund an AI buildout that could reach $70 billion this year — nearly triple its 2024 borrowing and the clearest signal yet that the AI race is now financed with debt.
-
Industry EN兩年從 108 億到 296 億美元：字節跳動史上最大聯貸，一個數字看懂 AI 的「舉債競賽」時代
字節跳動向近 30 家銀行簽下 296 億美元聯貸，資助今年最高可達 700 億美元的 AI 基礎建設——是 2024 年借款的近三倍，也是 AI 軍備競賽進入「槓桿時代」最清晰的訊號。
-
Research ENOne Model Finished the Hack: Booz Allen's Cyber Weapon Index Ranks 18 AIs as Attackers — and a Cheap Harness Erases the Ranking
Booz Allen put 18 US and Chinese AI models against a live corporate network as autonomous attackers. Only Claude Mythos completed the full kill chain — then 15th-place Claude Sonnet 5 matched it once given an attack harness.
-
Research 中只有一個模型完成入侵：Booz Allen「網路武器指數」讓 18 個 AI 當駭客實測，一套便宜工具就讓排名失效
Booz Allen 將 18 個美中新模型放到真實企業網路中當自主攻擊者，只有 Claude Mythos 完整走完攻擊殺傷鏈——但第 15 名的 Sonnet 5 加上攻擊框架後就能追平領先者。
-
Policy EN49 Projects Frozen, 117 in Limbo: Thailand Hits Pause on Its Data Center Boom
Thailand has suspended construction on 49 data center projects and stalled approvals for 117 more, giving agencies one month to draft environmental, economic, and power-grid rules for an industry growing faster than its oversight.
-
Policy 中49 項工程凍結、117 件卡關：泰國按下資料中心熱潮的暫停鍵
泰國宣布暫停 49 個資料中心建案、凍結另外 117 件的審批，給各部會一個月時間為這個成長速度遠超監管能量的產業制訂環境、經濟與電網規範。
-
Industry EN160,000 Huawei Chips for DeepSeek: Inside the Order That Could Reshape China's Compute Endgame
Bloomberg reports DeepSeek will deploy at least 160,000 Huawei Ascend 950DT accelerators at its 1GW Inner Mongolia data center — potentially the largest known Ascend cluster and a concrete step toward replacing Nvidia in China.
-
Industry 中16 萬顆華為晶片訂單：DeepSeek 的算力豪賭如何改寫中國 AI 佈局
Bloomberg 報導 DeepSeek 計畫在內蒙古 1GW 資料中心部署至少 16 萬顆華為 Ascend 950DT 加速器，可望成為已知最大規模的昇騰叢集，也是中國取代 Nvidia 算力最具體的一步。
-
Research ENOne Agentic Task, 10,000x the Footprint: Vals AI Puts Hard Numbers on AI's Energy Bill
Independent benchmarker Vals AI measured electricity, carbon, and water across 16 open-weight models and found long 'thinking' agentic tasks carry up to 10,000x the environmental impact of a simple query.
-
Research 中一次 Agentic 任務，一萬倍足跡：Vals AI 用實測數據揭開 AI 的能源帳單
獨立評測機構 Vals AI 對 16 個開放權重模型量測電力、碳排與水耗，發現長時間「思考」的 agentic 任務，環境衝擊最高可達簡單問答的 10,000 倍。
-
Models ENSaudi Arabia's HUMAIN Bets on China: humain-m3 Is a 428B-Parameter Arabic Frontier Built on MiniMax
HUMAIN's humain-m3 — a 428B-parameter Arabic MoE adapted from China's MiniMax-M3 — averages 89.37% across seven Arabic benchmarks, beating GPT-5.6 SOL and Opus 5 in the company's own tests.
-
Models 中沙烏地 HUMAIN 押注中國：humain-m3 是建立在 MiniMax 之上的 428B 參數阿拉伯語前沿模型
HUMAIN 的 humain-m3 是以中國 MiniMax-M3 為基礎改造的 428B 參數阿拉伯語 MoE 模型，在七項阿拉伯語基準測試平均拿下 89.37%，於公司自測中擊敗 GPT-5.6 SOL 與 Opus 5。
-
Research EN4.5 Billion TikTok Records, 289GB, Three Weeks: The Private-API Scrape That Just Landed on Hugging Face
An anonymous researcher scraped 4.5 billion TikTok video records through the app's private Android API in three weeks and published all 289GB on Hugging Face — no login, no account, just forged devices and reverse-engineered signatures.
-
Research 中45 億筆 TikTok 資料、289GB、三週完成：一場繞過 App 私有 API 的爬取行動登陸 Hugging Face
一位獨立研究者透過 TikTok App 的私有 Android API，在三週內爬取 45 億筆影片紀錄，並將 289GB 資料集完整公開在 Hugging Face——全程無帳號、無登入，靠的是偽造裝置與逆向簽名。
-
Industry ENOne Category, Half the Exports: South Korea's Chip Sales Triple to a Record $46.65 Billion
AI infrastructure demand from Google and Amazon pushed South Korea's August semiconductor exports up 209% to a record $46.65 billion — 47.5% of everything the country sold abroad.
-
Industry 中單一品項撐起半壁出口：南韓晶片出口暴增 209% 至 466.5 億美元歷史新高
Google 與 Amazon 的 AI 基礎設施需求，推動南韓 8 月半導體出口年增 209% 至 466.5 億美元歷史新高——佔該國當月總出口的 47.5%。
-
Models ENFrom Fourth to First in One Week: How Alibaba's Qwen3.8-Max-0902 Snapshot Conquered CodeArena
Alibaba's date-stamped Qwen3.8-Max-0902 update jumped from 1,669 to 1,691 on CodeArena: WebDev, dethroning Claude Opus 5 — not with a new architecture, but with one targeted RL post-training pass on coding and 'cowork' agent trajectories.
-
Industry ENByteDance's $29.6 Billion Loan: The Biggest AI Debt Deal Yet Shows Banks Can't Stop Lending
TikTok's parent upsized its record offshore loan from $20B to $29.6B after banks piled in $30B+ of orders — 68bps over SOFR, the cheapest AI money on the market — to bankroll a $70B capex year.
-
Industry 中字節跳動 296 億美元貸款落地：AI 史上最大債務融資，銀行團搶單搶到降息
TikTok 母公司將離岸聯貸從 200 億美元上調至 296 億美元，銀行認購超額突破 300 億——SOFR 加 68 個基點的定價，是市面上最便宜的 AI 資金，背後是高達 700 億美元的年度資本支出。
-
Industry ENMoonshot AI Quietly Files for Hong Kong IPO as Kimi K3 Demand Chases a $50 Billion Valuation
The Kimi maker has confidentially filed to list in Hong Kong, with a pre-IPO round in talks at roughly $50B — up from $35B in July — after Kimi K3 pushed ARR past $300M.
-
Industry 中Kimi 背後的月之暗面低調遞交香港 IPO 申請：Kimi K3 需求推升 500 億美元估值談判
Kimi 開發商月之暗面已保密遞交香港上市申請，據報導正洽談以約 500 億美元投前估值的 Pre-IPO 輪融資，較 7 月的 350 億美元再攀高峰。
-
Models ENQwen3.8-Max-0902 Debuts at #1 on Code Arena WebDev: Alibaba's Coding & Cowork Refresh Dethrones Claude Opus 5
Alibaba's post-trained Qwen3.8-Max-0902 jumped from 1,669 to 1,691 points to take the top spot on Code Arena's WebDev leaderboard, edging out Claude Opus 5 (Max) and Kimi K3 Max at a blended $5 per million tokens.
-
Models 中Qwen3.8-Max-0902 空降 Code Arena WebDev 榜首：阿里「編碼與協作」強化版以 1,691 分擊退 Claude Opus 5
阿里巴巴針對 Coding & Cowork 後訓練的 Qwen3.8-Max-0902，在 Code Arena WebDev 榜單從 1,669 躍升至 1,691 分奪冠，超越 Claude Opus 5（Max）與 Kimi K3 Max，混合單價僅約每百萬 token 5 美元。
-
Research ENAlibaba's Amap Team Open-Sources DreamX-Creator: A 7B Model That Generates Video and Sound Together at 2K
DreamX-Creator 1.0 jointly denoises audio and video streams in a single 7B model, adds RL with multimodal feedback, and refines output to 2K in one denoising step per chunk.
-
Research 中阿里巴巴高德團隊開源 DreamX-Creator：7B 參數模型一次生成 2K 影音，畫面與聲音不再是兩回事
DreamX-Creator 1.0 以單一 7B 模型同時去噪生成音訊與視訊串流，搭配多模態強化學習與每片段一步去噪的 2K 精煉管線。
-
Models ENDeepSeek Open-Sources Its First Multimodal Agent: V4-Flash-Vision-Exp Weights Go Public Under MIT
DeepSeek published the full 305B-parameter DeepSeek-V4-Flash-Vision-Exp checkpoint to Hugging Face under an MIT license — native FP8 weights, a 1M-token context, DSpark speculative decoding, and agent scores within a point of Opus-4.8 on five benchmarks.
-
Models 中DeepSeek 開源首款多模態代理模型：V4-Flash-Vision-Exp 權重以 MIT 授權公開
DeepSeek 將完整的 305B 參數 DeepSeek-V4-Flash-Vision-Exp 檢查點以 MIT 授權上傳至 Hugging Face——原生 FP8 權重、百萬 token 上下文、DSpark 投機解碼，並在半數多模態代理基準上追平甚至超越 Opus-4.8。
-
Industry ENZ.ai's Revenue Quintupled to $142 Million in H1 — and Its Losses Are Finally Shrinking
The GLM maker's first-half sales jumped ~400% on an API surge, yet the stock fell: a look inside China's AI price war economics.
-
Industry 中Z.ai 上半年營收暴增至 1.42 億美元——虧損終於開始收斂
GLM 模型開發商上半年營收年增約 400%，受 API 業務暴增推動，但股價反應冷淡：深入解析中國 AI 價格戰的經濟學。
-
Models ENMoonshot Retires Kimi K2.5, Bets the Company on K3 at 5x the Price
Moonshot AI completed the retirement of Kimi K2.5 and moonshot-v1 on August 31, leaving the 2.8T-parameter Kimi K3 as its only flagship — at roughly 5x the per-token price of its predecessor, a deliberate reversal of China's year-long price war.
-
Models 中Moonshot 退役 Kimi K2.5，以貴五倍的 K3 押上全部身家
Moonshot AI 於 8 月 31 日完成 Kimi K2.5 與 moonshot-v1 的退役，2.8 兆參數的 Kimi K3 成為唯一現役旗艦——每 token 价格約為前代的五倍，等於親手終結了中國 AI 界長達兩年的價格戰。
-
Research ENByteDance's Lucida Turns Messy Room Videos Into Editable 3D Scenes for Robots
ByteDance Seed's Lucida pipeline parses cluttered indoor video into per-instance scene graphs, generates an asset per object, and lets a VLM policy drive the 3D editor's own gizmo handles until it decides placement is done — posting a 69% mAP gain on R2S-Scene.
-
Research 中ByteDance Lucida：把雜亂房間影片變成機器人可用的可編輯 3D 場景
ByteDance Seed 的 Lucida 管線把雜亂室內影片解析成逐物件場景圖、為每個物件生成資產，再由 VLM 策略 GizmoAct 直接操作 3D 編輯器的 gizmo 手柄決定擺放完成與否——在 R2S-Scene 上繳出 69% 的 mAP 增益。
-
Industry ENDeepSeek Nears a $7.4 Billion Round at a $74 Billion Valuation — Its Second Mega-Raise in Three Months
Fresh off its first-ever outside round in June, DeepSeek is closing in on roughly 50 billion yuan more at a ~500 billion yuan pre-money valuation, with Monolith, Shixiang, CATL and local-government funds lining up ahead of a possible 2027 Shanghai listing.
-
Industry 中DeepSeek 將以 740 億美元估值完成 74 億美元融資——三個月內第二筆巨額注資
繼 6 月首度對外募資後，DeepSeek 即將再籌約 500 億人民幣，投前估值約 5,000 億人民幣，Monolith、勢詳、寧德時代與地方政府基金均已表態參與，為 2027 年上海上市鋪路。
-
Industry ENZhipu's Revenue Quintuples as Z.AI Posts First Results as a Public AI Lab
Zhipu (Z.AI) reported H1 revenue of 953.9 million yuan, up 400%, while narrowing its net loss — the first big earnings test for China's open-source model champion.
-
Industry 中智谱营收暴增五倍：Z.AI 交出上市後首份公開財報
智谱（Z.AI，2513.HK）公布上半年營收 9.539 億人民幣、年增約 400%，淨虧損同步收窄——中國開源模型旗手迎來上市後第一場大型財報考驗。
-
Models ENTencent Open-Sources Hy4 Preview: A 770B MoE Flagship With 1M-Token Context
Tencent's Hunyuan team releases Hy4 preview, a 770B-parameter open-weight MoE with 49B active parameters, a 1M-token context window, and coding results that rival GLM-5.3 and Kimi K3.
-
Models 中騰訊開源 Hy4 Preview：770B 參數 MoE 旗艦模型，支援百萬 token 上下文
騰訊混元團隊發布 Hy4 preview，770B 總參數的開放權重 MoE 模型，每 token 僅啟動 49B 參數，支援 100 萬 token 上下文，編碼表現超越 GLM-5.3 與 Kimi K3。
-
Industry ENChina's CXMT Begins First HBM3E Production, Cracking Open the AI Memory Bottleneck
ChangXin Memory Technologies has started small-batch production of HBM3E, China's first domestically made advanced AI memory — a generation behind the leaders, but a strategic breakthrough against export controls.
-
Industry 中中國 CXMT 首度投產 HBM3E，撬開 AI 記憶體瓶頸
長鑫存储（CXMT）已開始小批量生產 HBM3E，這是中國首度自產先進 AI 記憶體——雖落後龍頭一個世代，卻是對抗出口管制的戰略性突破。
-
Industry ENTaiwan Raids Nvidia Supplier Unimicron Over China-Parts Relabeling
Prosecutors raided Unimicron's Taoyuan headquarters over allegations it relabeled China-made PCBs as Taiwanese, hitting shares 10% and exposing origin-fraud risk at the heart of the AI supply chain.
-
Industry 中台灣檢方搜索輝達供應商欣興電子 揭中國製零件「洗產地」疑雲
桃園檢方搜索欣興電子總部，質疑其將中國製 PCB 回台重新標示為台灣製，股價重挫約 10%，也讓 AI 供應鏈的產地詐欺風險一次浮上檯面。
-
Models ENTencent Open-Sources Hy4 Preview: A 770B-Parameter MoE Frontier Model with 1M-Token Context
Tencent's Hunyuan team releases Hy4 preview under Apache 2.0: 770B total parameters, 49B active per token, a 1M-token context window, and benchmark scores that edge out GLM-5.3 and Kimi K3 on real-world engineering tasks.
-
Models 中騰訊開源 Hy4 preview：770B 參數 MoE 前沿模型、百萬 token 上下文
騰訊混元團隊以 Apache 2.0 授權釋出 Hy4 preview：總參數 770B、每 token 啟用 49B、上下文超過 100 萬 token，在盲測工程任務上小勝 GLM-5.3 與 Kimi K3，直攻開源前沿。
-
Industry ENDeepSeek Nears $7.4B Round at $74B Valuation as 2027 STAR Market IPO Takes Shape
DeepSeek is closing a fresh ~50 billion yuan ($7.4B) round at a ~500 billion yuan (~$74B) pre-money valuation, targeting an end-of-August close that paves the way for a possible 2026 IPO filing and a 2027 Shanghai STAR Market debut.
-
Industry 中DeepSeek 逼近以 740 億美元估值完成 74 億美元融資，2027 科創板 IPO 藍圖成形
DeepSeek 正接近完成約 500 億人民幣（74 億美元）的新一輪融資， pre-money 估值約 5,000 億人民幣（約 740 億美元），目標 8 月底完成交割，為可能於 2026 年底遞件、2027 年在上海科創板掛牌鋪路。
-
Research ENChatbots Debunked Foreign Propaganda 75% of the Time — and Beat Search Engines in NPR's Test
NPR and NewsGuard posed 30 questions built from 15 false narratives pushed by Russia, China and Iran to six chatbots and four search engines. Chatbots debunked about three-quarters of them and failed less often than search — but AI summaries sitting on top of search results fared worst of all.
-
Research 中NPR 實測：聊天機器人破解外國宣傳的成功率達 75%，表現勝過搜尋引擎
NPR 與 NewsGuard 以俄羅斯、中國、伊朗散布的 15 個假敘事設計出 30 道問題，測試六款聊天機器人與四大搜尋引擎。聊天機器人平均破解約四分之三的假敘事，失敗率低於傳統搜尋——但疊在搜尋結果頂端的 AI 摘要表現最差。
-
Policy ENUS Drafts Rule to Close the Remote GPU Rental Loophole — but Lawyers Say BIS Lacks the Authority
Commerce is preparing a September rule to block Chinese AI firms from renting Nvidia GPU time through Thai and Singapore data centers, but export-control lawyers say BIS has no statutory authority over remote compute access.
-
Policy 中美國擬新規堵住遠端 GPU 租用漏洞——但律師說商務部根本沒有執法權
美國商務部正起草九月新規，要阻止中國 AI 公司透過泰國、新加坡資料中心租用 Nvidia GPU 算力，但出口管制律師普遍認為 BIS 對遠端算力存取缺乏法定授權。
-
Industry ENChina's Robot Olympics Ends: Tiangong Ultra Runs 100m in 8.64 Seconds as AGIBOT Tops the Medal Table
The second World Humanoid Robot Games closed in Beijing with a record 8.64-second 100m sprint, 46 medals for AGIBOT, and 2,056 competing robots — a five-day snapshot of how fast humanoid hardware is actually maturing.
-
Industry 中中國「機器人奧運」落幕：天工 Ultra 百米跑出 8.64 秒，AGIBOT 稱霸獎牌榜
第二屆世界人形機器人運動會在北京閉幕：天工 Ultra 以 8.64 秒奪下百米決賽、AGIBOT 拿下 46 面獎牌、2,056 台機器人同場競技——五天賽事是人形硬體成熟速度的最佳縮影。
-
Models ENAugust 31 Is AI's Deadline Day: Three Model Sunsets Hit at Once
Tomorrow, Claude Sonnet 5's $2/$10 intro pricing expires, GPT-5.4 leaves Codex for ChatGPT sign-in users, and Moonshot fully sunsets kimi-k2.5 and moonshot-v1 — a rare triple deadline that will reprice and reroute a huge slice of production AI traffic in a single day.
-
Models 中8 月 31 日是 AI 的截止日：三個模型日落同一天降臨
明天，Claude Sonnet 5 的 2 美元／10 美元促銷價到期，GPT-5.4 從 Codex 的 ChatGPT 登入環境退場，Moonshot 則全面下架 kimi-k2.5 與 moonshot-v1——三個獨立期限罕见地落在同一天，將在 24 小時內重新定價並改道大量生產環境的 AI 流量。
-
Industry ENNvidia Wants to Run the World's Robots — and China Is Its Best Customer
A WSJ investigation finds Chinese robotics firms quietly became one of the biggest buyers of Nvidia's physical-AI stack — a business already at $10B a year that Jensen Huang expects to grow 10-fold within a decade.
-
Industry 中輝達想驅動全世界的機器人——而最大客戶是中國
《華爾街日報》調查發現，中國機器人公司已悄悄成為輝達「實體 AI」技術棧的最大買家之一——這門生意年營收已達 100 億美元，黃仁勳預期十年內成長十倍。
-
Tools ENvLLM 0.28.0 Lands Decode Context Parallel, DFlash2 Speculative Decoding and Tiered Disk KV Offload
The vLLM project's latest release packs 584 commits from 270 contributors: Decode Context Parallel for Kimi-K3, end-to-end sparse MLA for DeepSeek V4, DFlash2 speculation, disk-tier KV offloading, and doubled batch defaults.
-
Tools 中vLLM 0.28.0 釋出：Decode Context Parallel、DFlash2 投機解碼與磁碟層 KV 卸載全面到位
vLLM 最新版本集結 270 位貢獻者的 584 項提交：Kimi-K3 的 Decode Context Parallel、DeepSeek V4 端到端稀疏 MLA、DFlash2 投機解碼、磁碟層 KV 快取卸載，以及翻倍的批次預設值。
-
Policy ENCXMT Sues the Pentagon to Escape the 'Chinese Military Company' Blacklist
China's top DRAM maker ChangXin Memory Technologies has sued the U.S. Defense Department over its Section 1260H 'Chinese military company' designation — armed with an unusual piece of evidence: the Pentagon once told CXMT it was off the list, then changed its mind within hours.
-
Policy 中CXMT 控告美國國防部：中國記憶體龍頭全力反擊「中國軍事企業」黑名單
中國最大 DRAM 廠長鑫存儲（CXMT）向美國哥倫比亞特區聯邦地方法院提起訴訟，挑戰五角大廈依《國防授權法》第 1260H 條將其列入「中國軍事企業」清單的決定——而且握有一項不尋常的證據：五角大廈曾通知 CXMT 已將其移出名單，卻在數小時內反悔。
-
Industry ENCXMT Beats Samsung and SK Hynix to LPDDR6: Chinese Memory Debuts on Xiaomi's 18 Fold
China's CXMT has begun mass-producing LPDDR6 mobile memory and will debut it inside Xiaomi's 18 Fold flagship in September — the first time a Chinese DRAM maker ships next-generation memory in a premium phone ahead of Samsung, SK Hynix, and Micron.
-
Industry 中CXMT 搶先三星、SK 海力士量產 LPDDR6：中國記憶體首度登上小米 18 Fold 旗艦
中國最大記憶體廠長鑫存儲（CXMT）宣布 LPDDR6 進入量產，並將於 9 月在小米 18 Fold 旗艦摺疊手機首發——這是中國 DRAM 廠商首次與三星、SK 海力士、美光同步、甚至搶先在新一代行動記憶體上出貨。
-
Models ENTencent Open-Sources Hy4 Preview: A 770B MoE Flagship With 1M-Token Context
Tencent's Hy Team has released Hy4 preview, a 770B-parameter open-weight MoE model with 49B active parameters, a 1M-token context window, and benchmark wins over GLM-5.3 and Kimi K3.
-
Models 中騰訊開源 Hy4 preview：770B 參數 MoE 旗艦模型，支援百萬 token 上下文
騰訊 Hy 團隊開源 Hy4 preview：770B 總參數、每 token 僅啟動 49B 的 MoE 旗艦模型，具備百萬 token 上下文視窗，評測成績超越 GLM-5.3 與 Kimi K3。
-
Policy ENInside Taiwan's B300 Smuggling Case: Nine Indicted, One of Them From Nvidia
Taiwan's Keelung District Prosecutors have indicted nine people — including an Nvidia distribution manager and two former Supermicro employees — for falsifying paperwork on 130 B300 AI servers, 74 of which reached Chinese buyers via Indonesia, Japan and Hong Kong.
-
Policy 中台灣 B300 走私案起訴九人：其中一人來自 NVIDIA
台灣基隆地檢署起訴九人，包括一名 NVIDIA 通路經理與兩名前 Supermicro 員工，涉嫌偽造 130 台 B300 AI 伺服器文件，其中 74 台經印尼、日本與香港轉運後流入中國買家手中。
-
Models ENThe $10 Billion Clause: Z.ai's GLM-5.3 License Puts a Price Tag on Trust
GLM-5.3's weights are finally public — under a bespoke license that forces any $10B+ Model-as-a-Service provider through Z.ai's security review. As US labs reel from a summer of AI security scares, China is pitching openness itself as the safer bet.
-
Models 中100 億美元條款：Z.ai 的 GLM-5.3 授權為「信任」標上價格
GLM-5.3 權重終於公開，但採用了一份量身打造的授權：年營收超過 100 億美元的模型服務業者，必須先通過 Z.ai 的安全審查。在美國實驗室飽受一連串 AI 安全事件衝擊之際，中國正把「開放」本身包裝成更安全的選擇。
-
Models ENQwen3.8-Flash-Next: Alibaba Open-Sources the First Glimpse of Qwen4's Architecture
Alibaba's Qwen team open-weights a 125B-parameter MoE that activates just 6B per token, pairs 1M-token context with a hybrid GDN + QSA attention stack, and beats Claude Opus 4.6 Max on agentic coding benchmarks.
-
Models 中Qwen3.8-Flash-Next：阿里巴巴開源 Qwen4 架構的首波預覽
阿里巴巴 Qwen 團隊開源一款 125B 參數的 MoE 模型，每 token 僅啟動 6B，兼顧百萬級上下文與 GDN + QSA 混合注意力，在代理式編程基準上超越 Claude Opus 4.6 Max。
-
Models ENFive Open-Weight Models in Nine Days: The Capability Premium Just Collapsed
Between August 21 and 29, five Chinese AI labs shipped open-weight models with 1M-token context at budget prices — and OpenAI answered with a 20% price cut. The frontier is now a price war.
-
Models 中九天五款開源權重模型：能力溢價正在崩塌
8 月 21 日至 29 日，五家中國 AI 實驗室接連推出具備百萬 token 上下文的開源權重模型，價格卻壓在低價帶——OpenAI 的回應是降價 20%。前沿模型市場正式進入價格戰。
-
Models ENTencent Open-Sources Hy4 Preview: A 770B MoE Flagship Built for Real Work
Tencent's Hunyuan team releases Hy4 preview, a 770B-parameter MoE model with 49B active parameters, 1M-token context and an early recursive self-improvement loop.
-
Models 中騰訊開源 Hy4 Preview：770B 參數 MoE 旗艦模型，還學會了自我改進
騰訊混元團隊發布 Hy4 preview：770B 總參數、49B 激活參數的 MoE 旗艦模型，具備百萬 token 上下文與早期遞迴自我改進迴路，採 Apache 2.0 開源。
-
Industry ENMiniMax Triples Alibaba Cloud Spending Ceiling to US$1.2 Billion as Compute Hunger Bites
New HKEX filings show MiniMax has raised its three-year Alibaba Cloud purchase cap 220% to US$1.2 billion after burning through two-thirds of its annual budget by June — a concrete picture of what surging AI inference demand costs in China's cloud market.
-
Industry 中運算需求暴增，MiniMax 將阿里雲採購上限調高三倍至 12 億美元
港交所文件顯示，MiniMax 因上半年就用掉三分之二的年度雲端預算，將與阿里雲的三年採購上限調高 220% 至 12 億美元，具體呈現了中國 AI 推論需求爆炸的真實成本。
-
Models ENDeepSeek's V4-Flash-Vision-Exp Sees for Pennies: The 384-Token Gamble Reshaping Agent Economics
DeepSeek's experimental multimodal model matches V4-Flash text performance at identical prices, caps every image at 384 tokens, and trails Opus-4.8 by single points on multimodal agent benchmarks — while coding harnesses burn billions of tokens through it in days.
-
Models 中DeepSeek V4-Flash-Vision-Exp：每張圖 384 Token 的視覺賭注，正在改寫 Agent 經濟學
DeepSeek 的實驗性多模態模型以與 V4-Flash 完全相同的價格提供視覺能力，每張圖片固定計費 384 token，多模態 Agent 基準逼近 Opus-4.8——上線三天就被編碼代理燒掉 250 億 token。
-
Industry ENXPeng's IRON Humanoid Unit Raises $900 Million in China's Largest Physical AI Funding Round
XPeng's robotics unit Dogotix has raised over $900 million at a $6.3 billion valuation to mass-produce its IRON humanoid robot — the largest single-round private financing in China's physical AI industry, led by IDG Capital with Tencent and Alibaba as strategic investors.
-
Industry 中小鵬 IRON 人形機器人部門融資 9 億美元,創中國實體 AI 最大單輪紀錄
小鵬汽車旗下機器人業務 Dogotix 以超過 63 億美元估值融資逾 9 億美元,用於 IRON 人形機器人量產——這是中國實體 AI 產業迄今最大單輪私募融資,由 IDG 資本領投,騰訊與阿里巴巴戰略投資。
-
Policy ENFake Think Tank, Real AI: Inside OpenAI's Takedown of Russia's 'Burke Institute' Influence Operation
OpenAI banned a cluster of Russia-origin ChatGPT accounts that promoted the International Burke Institute — a fake Israel-based think tank with a plagiarized article archive and a proprietary 'Sovereignty Index' engineered to flatter Russia. A close look at how AI became the scaffolding, and the undoing, of a manufactured authority.
-
Policy 中假智庫、真 AI：拆解 OpenAI 粉碎俄羅斯「Burke 研究所」影響力行動的始末
OpenAI 封鎖了一群源自俄羅斯的 ChatGPT 帳號——它們推廣自稱位於以色列的「國際 Burke 研究所」：一個抄襲論文、發明「主權指數」來吹捧俄羅斯的假智庫。本文深入解析 AI 如何成為這場「製造權威」行動的鷹架，又如何成為它敗露的關鍵。
-
Models ENTencent Open-Sources Hy4 Preview: 770B MoE Flagship With 1M Context and a 31.8% Self-Tuning Speedup
Tencent Hunyuan's new open-weight flagship packs 770B total / 49B active parameters, a 1M-token window, and an early recursive self-improvement loop that lifted its own inference throughput 31.8%.
-
Models 中騰訊開源 Hy4 Preview：770B MoE 旗艦、百萬 token 上下文，還親手把自己的推理吞吐提速 31.8%
騰訊混元新一代開源旗艦 Hy4 preview 擁有 770B 總參數、49B 激活參數與超過 1M token 的上下文視窗，並首度建立遞迴自我改進迴路，自行優化推論系統帶來 31.8% 吞吐提升。
-
Models ENZ.ai's GLM-5.3-Flash: Frontier Multimodal Intelligence at One-Tenth the Cost
Z.ai open-sources GLM-5.3-Flash, a 320B-parameter hybrid-attention multimodal model that matches Claude Opus 4.8 on coding at roughly a tenth of the price — with 1M-token context, MIT license, and inference served at scale on Chinese AI chips.
-
Models 中Z.ai 的 GLM-5.3-Flash：十分之一成本的前沿多模態模型
Z.ai 開源 GLM-5.3-Flash：320B 參數混合注意力多模態模型，程式能力接近 Claude Opus 4.8、價格僅約十分之一，支援百萬 token 上下文、MIT 授權，並大規模部署於中國自研 AI 晶片上。
-
Models ENGLM-5.3 Open Weights Land: Z.ai's Exploit-Hunting Flagship Finally Goes Public
After a two-week safety review, Z.ai's GLM-5.3 open weights are scheduled to hit Hugging Face today — and GLM-5.3-Flash's MIT-licensed weights are already live, turning every self-reported benchmark claim into something anyone can verify.
-
Models 中GLM-5.3 開放權重上線：Z.ai 那個會找漏洞的旗艦模型終於公開
歷經兩週安全審查，Z.ai 的 GLM-5.3 開放權重預計今（28）日登上 Hugging Face，而 MIT 授權的 GLM-5.3-Flash 權重已經上線——所有自報的基準分數，從今天起人人都可以親自驗證。
-
Industry ENChina Now Runs Over 2 Million Factory Robots — and Half the World Makes Them
A BBC on-the-ground report finds more than 2 million industrial robots at work in Chinese factories, fueled by a shrinking workforce and a $20 billion state plan.
-
Industry 中中國工廠機器人突破 200 萬台——全球一半機器人都是它造的
BBC 實地報導揭露：中國工廠在役工業機器人已超過 200 萬台居全球之冠，背後是人口萎縮與 200 億美元的國家自動化戰略。
-
Industry ENHuawei Widens Its AI Pharma Push: From Compound Screening to the Clinic
Huawei's healthcare president William Zhang says the company will expand AI cooperation with Chinese drugmakers into drug development and clinical practice — riding its Ascend and Kunpeng chips into a market where Nvidia has already partnered with Eli Lilly and Novo Nordisk.
-
Industry 中華為擴大 AI製藥布局：從分子篩選一路推進臨床
華為醫療事業總裁張強表示，將把與中國藥企的 AI 合作從早期研究擴展到藥物開發與臨床實踐——憑藉昇騰與鯤鵬晶片，切入輝達已與禮來、諾和諾德結盟的 AI 製藥市場。
-
Industry ENTIME's 2026 TIME100 AI List Maps the New Power Structure of Artificial Intelligence
TIME's latest TIME100 AI collection — published today — crowns the OpenAI trio, Anthropic's founders, and an infrastructure-heavy cast from Broadcom's Hock Tan to Micron's Sanjay Mehrotra, alongside a deep bench of Chinese AI leaders, agent-economy founders, and Hollywood creators fighting for AI rights.
-
Industry 中TIME 2026「AI 百大影響力人物」出炉：一幅新的 AI 權力地圖
TIME 今日發布 2026 年 TIME100 AI 名單：OpenAI 三巨頭、Anthropic 創辦人兄弟檔，加上從博通 Hock Tan 到美光 Sanjay Mehrotra 的基礎設施豪門、一字排開的中國 AI 領袖、Agent 新創創辦人，以及為創作者權益發聲的好萊塢名人。
-
Industry ENMoonshot AI Wants 30% of Hyperscaler Revenue to Host Kimi K3 — and It Might Have the Leverage
Reuters reports that China's Moonshot AI is in early talks with Microsoft, Amazon, and Google to host Kimi K3 on their clouds — demanding up to 30% of K3-related revenue, a deal that would invert the app-store model and test how much US hyperscalers need frontier open-weight models.
-
Industry 中Moonshot AI 要求三大雲端巨頭上繳 30% 營收才能託管 Kimi K3——而它可能真有談判籌碼
路透社獨家報導，中國 Moonshot AI 正與微軟、亞馬遜、Google 早期洽談，要求在 Azure、AWS、Google Cloud 上託管 Kimi K3 的營收分成最高可達 30%——這樁交易若成，將是首件中國 AI 實驗室與美國雲端巨頭的營收共享協議，並徹底顛覆 app store 式的平台抽成邏輯。
-
Models ENOx Alpha Unmasked: Z.ai's Stealth Model Is GLM-5.3-Flash, a 320B Open-Weight Multimodal Contender
The anonymous model that topped OpenRouter's usage charts turned out to be Z.ai's GLM-5.3-Flash — a 320B-A18B natively multimodal MoE with 1M context and MIT-licensed weights at a tenth of frontier pricing.
-
Models 中Ox Alpha 揭曉真身：Z.ai 的匿名模型就是 GLM-5.3-Flash——320B 開源多模態強權
登上 OpenRouter 用量冠軍的匿名模型「Ox Alpha」，證實是 Z.ai（智譜）的 GLM-5.3-Flash——320B-A18B 原生多模態 MoE、百萬 token 上下文、MIT 授權開放權重，價格僅前線模型的十分之一。
-
Policy ENTaiwan Indicts Nine in Nvidia B300 Server Smuggling Ring — 74 Racks Reached China
Taiwanese prosecutors charged nine people — including an Nvidia senior manager and two Supermicro employees — with forging documents to smuggle 130 B300 AI servers to China; 74 got through via Indonesia, Japan, and Hong Kong before customs caught the rest.
-
Policy 中台灣起訴九人:輝達 B300 伺服器走私網——74 台已流入中國
台灣檢方依背信與偽造文書罪起訴九人,包括輝達台灣辦公室一名高級經理與兩名美超微員工,指控他們偽造文件將 130 台 B300 AI 伺服器走私至中國;其中 74 台經印尼、日本與香港成功交付,其餘遭海關攔截。
-
Models ENQwen3.8-Flash-Next: Alibaba Opens the Door on the Qwen4 Architecture
Alibaba's Qwen team releases a 125B-parameter open-weight preview of the Qwen4 architecture — hybrid linear attention, a 51B-parameter n-gram embedding, and 6B active parameters per token.
-
Models 中Qwen3.8-Flash-Next：阿里巴巴提前揭開 Qwen4 架構的面紗
阿里巴巴 Qwen 團隊發布 125B 參數的開放權重模型，預覽 Qwen4 架構：混合線性注意力、51B 參數的 n-gram 嵌入層，以及每 token 僅 6B 活躍參數。
-
Models ENMystery Solved: Ox Alpha Was GLM-5.3-Flash All Along — 320B Open Weights for $0.15/M Tokens
Z.ai unmasked its anonymous stealth model: GLM-5.3-Flash is a 320B-A18B natively multimodal MoE with 1M context and MIT-licensed weights at $0.15/$0.50 per million tokens — and the entire preview ran on Chinese-made AI chips.
-
Models 中謎底揭曉：Ox Alpha 就是 GLM-5.3-Flash——320B 開放權重、每百萬 token 只要 0.15 美元
Z.ai 揭曉匿名 stealth 模型的真實身分：GLM-5.3-Flash 是 320B-A18B 原生多模態 MoE，具備 1M context 與 MIT 授權權重，API 定價每百萬 token 輸入 0.15／輸出 0.50 美元——而且整個預覽期間都跑在中國自產 AI 晶片上。
-
Models ENOx Alpha Revealed: Z.ai's GLM-5.3-Flash Ships Open Weights, Frontier Scores, and a Chinese-Chip Serving Stack
Z.ai confirmed the stealth model Ox Alpha is GLM-5.3-Flash: a 320B-parameter MIT-licensed multimodal model with near-frontier coding scores, a 1M-token context window, aggressive pricing, and inference served on tens of thousands of domestic Chinese AI chips.
-
Models 中Ox Alpha 揭曉：Z.ai 的 GLM-5.3-Flash 開放權重、前線級分數與中國晶片推理叢集
Z.ai 證實 stealth 神秘模型 Ox Alpha 就是 GLM-5.3-Flash：320B 參數、MIT 授權開放權重的原生多模態模型，具備百萬 token 上下文與逼近前線的編碼分數，且整個曝光週期的推理服務都跑在數萬顆中國國產 AI 晶片上。
-
Industry ENYMTC's $4.9 Billion IPO Filing Takes Aim at Samsung and SK Hynix
China's NAND champion YMTC has filed for a $4.9 billion Shanghai STAR Market listing through parent CCSH — and told IPO investors it intends to become the world's largest flash supplier by the end of 2027, overtaking Samsung and SK Hynix.
-
Industry 中YMTC 49 億美元 IPO 申請，劍指三星與 SK 海力士
中國 NAND 快閃記憶體龍頭長江存儲（YMTC）透過母公司長江存儲科技控股（CCSH）申請在上海科創板上市，募資 49 億美元，並向 IPO 投資人表明：目標在 2027 年底前超越三星與 SK 海力士，成為全球最大快閃記憶體供應商。
-
Industry ENMiniMax Nearly Quadruples Revenue in First Post-IPO Interim Results, as Enterprise API Demand Explodes 703%
MiniMax reported H1 2026 revenue of US$116.6 million, up 283% year over year and already exceeding all of 2025, with Open Platform enterprise revenue surging 703% as token consumption hit 20x its January level.
-
Industry 中MiniMax 上市後首份中期財報營收近翻四倍，企業 API 業務暴增 703%
MiniMax 公布 2026 上半年營收 1.166 億美元、年增 283%，已超越 2025 全年總和；Open Platform 企業業務暴增 703%，Token 消耗量更達 1 月的 20 倍。
-
Industry ENDeepSeek's Revenue Jumped Tenfold to $70.7M — and Its API Business Is More Profitable Than OpenAI's
Leaked financials show DeepSeek generated RMB 475M ($70.7M) in the first seven months of 2026, a tenfold jump, with API gross margin at a remarkable 82.9%.
-
Industry 中DeepSeek 營收暴增十倍至 7,070 萬美元——API 毛利率竟超越 OpenAI
外流財報顯示，DeepSeek 2026 年前七個月營收達 4.75 億人民幣（7,070 萬美元），年增十倍，API 毛利率更高達 82.9%。
- Industry EN
Two Million Robots and Counting: Inside China's Factory Automation Sprint Against Its Own Demographics
A BBC on-the-ground report finds more than two million industrial robots at work in Chinese factories — over half the world's supply is now built in China — as Beijing races to automate before an aging, shrinking workforce catches up with its economy.
- Industry 中
兩百萬台機器人還在增加:透視中國工廠自動化與人口老化的賽跑
BBC 實地報導發現,中國工廠裡已有超過兩百萬台工業機器人運轉——全球一半以上的工業機器人由中國製造——北京正與時間賽跑,希望在人口老化與萎縮拖垮經濟之前完成自動化轉型。
-
Industry ENEnflame, the Last of China's Four AI Chip Dragons, Opens Its $892 Million STAR Market IPO
Tencent-backed Enflame opens subscriptions September 2 for a 6 billion yuan Shanghai STAR Market listing — the final IPO among China's 'four little dragons' of AI silicon, and a referendum on Beijing's chip-independence push.
-
Industry 中燧原科技掛牌倒數：中國「AI 晶片四小龍」最後一員啟動 892 億台幣級 STAR 科創板 IPO
騰訊投資的燧原科技宣布 9 月 2 日開放上交所科創板 IPO 申購，募資約 60 億人民幣（8.92 億美元）——這是「四小龍」中最後掛牌的一家，也是對中國晶片自主之路的一場公開投票。
-
Models ENNobody Knows Who Built Ox Alpha: The Anonymous Model Beating GPT-5.6 at Coding
A frontier-class stealth model called Ox Alpha appeared free on OpenRouter on August 20 with a 1M-token context window, video input, and an 80% DeepSWE score that tops Claude Fable 5 and GPT-5.6 Sol — and its creator is staying anonymous.
-
Models 中沒人知道誰做了 Ox Alpha：擊敗 GPT-5.6 的匿名模型
一款名為 Ox Alpha 的前沿級隱身模型於 8 月 20 日免費現身 OpenRouter，具備百萬 token 上下文、視訊輸入，以及以 80% 的 DeepSWE 分數壓過 Claude Fable 5 與 GPT-5.6 Sol——而它的創造者選擇匿名。
-
Models ENAlibaba Previews the Qwen4 Architecture Early: Qwen3.8-Flash-Next Open Weights Drop August 26
Hours before the weights go live, Alibaba is shipping its Qwen4 architecture — GDN hybrid attention plus Qwen Sparse Attention — inside an open-weight multimodal MoE branded Qwen3.8-Flash-Next, rumored at 125B total / 6B active parameters.
-
Models 中阿里巴巴提前公開 Qwen4 架構：Qwen3.8-Flash-Next 開放權重 8 月 26 日登場
在權重上架前的幾小時，阿里巴巴將下一代 Qwen4 架構——GDN 混合注意力加上 Qwen Sparse Attention——包進開放權重的多模態 MoE 模型 Qwen3.8-Flash-Next 釋出，傳聞總參數 125B、激活參數 6B。
-
Industry ENOpenAI-Backed Harvey Builds Its First In-House Model on China's Kimi K3
Legal AI leader Harvey post-trained Harvey Tenet on Moonshot AI's open-weight Kimi K3 — the clearest sign yet that Chinese open models are becoming the foundation layer for Western enterprise AI.
-
Industry 中OpenAI 投資的 Harvey，第一款自研模型選擇了中國的 Kimi K3 作為基底
法律 AI 龍頭 Harvey 以月之暗面的開源權重模型 Kimi K3 為基底，後訓練出首款自研模型 Harvey Tenet——這是中國開源模型成為西方企業 AI 基礎層最明確的訊號。
-
Models ENMiniMax Open-Sources Music 3: Complete Five-Minute Songs From One Prompt
MiniMax Music 3 pairs an 8B Global LLM with a 0.6B Local LLM and a flow-matching synth stage to generate full five-minute songs — and the weights are on Hugging Face.
-
Models 中MiniMax 開源 Music 3：一次生成完整五分鐘歌曲
MiniMax Music 3 結合 8B 全域 LLM、0.6B 區域 LLM 與流匹配合成階段，可生成完整五分鐘歌曲——模型權重已上架 Hugging Face 供所有人下載。
-
Industry ENUnitree's 45% Post-IPO Slump Sparks Bubble Fears in China's Robotics Boom
Unitree shares have lost 45% since their 460% first-day surge in Shanghai, erasing roughly $30 billion and igniting a debate over froth, retail losses and China's IPO pricing system.
-
Industry 中宇樹科技IPO後暴跌45%，中國機器人熱潮引發泡沫疑慮
宇樹科技上市首日大漲460%後，五個交易日內市值蒸發約300億美元，引發市場對泡沫風險、散戶虧損與中國IPO定價機制缺陷的激烈討論。
-
Models ENOx Alpha: The Anonymous Frontier Model Nobody Can Trace
A stealth AI model with 1M-token context appeared free on OpenRouter and started beating GPT-5.6 and Claude on coding benchmarks. Nobody knows who built it — but the fingerprints are getting clearer.
-
Models 中Ox Alpha：沒有人能追溯的匿名前沿模型
一個擁有百萬 token 上下文視窗的隱身 AI 模型免費現身 OpenRouter，並在程式碼基準測試中擊敗 GPT-5.6 與 Claude。沒有人知道它是誰打造的——但指紋線索正越來越清晰。
-
Policy ENDeepSeek Is Now the Weapon of Choice for Chinese State Hackers — And It Doubled Their Attack Volume
TeamT5 says Chinese state-affiliated groups have more than doubled attack volume by wiring DeepSeek into reconnaissance, exploit writing, and lateral movement — drawn by weak cyber guardrails and rock-bottom costs.
-
Policy 中DeepSeek 成為中國駭客國家隊的新武器——攻擊量直接翻倍
台灣資安廠商 TeamT5 提出報告：中國官方背景駭客組織將 DeepSeek 導入偵察、漏洞攻擊與橫向移動流程後，攻擊量已翻倍——低廉成本與寬鬆防護欄是主因。
-
Meta ENXPeng's Robotics Arm Raises $900M at $6.3B Valuation — China's Largest Embodied AI Round Ever
XPeng's robotics business closed China's largest embodied-AI funding round — over $900M led by IDG Capital with Tencent and Alibaba backing — to push the IRON humanoid into mass production by end-2026.
-
Meta 中小鵬機器人業務募資 9 億美元、估值 63 億美元——中國具身智慧史上最大單輪融資
小鵬汽車機器人業務完成中國具身智慧產業最大單輪融資——IDG 資本領投超過 9 億美元，騰訊、阿里巴巴戰略加持，目標 2026 年底讓 IRON 人形機器人進入量產。
-
Policy ENTaiwan Indicts Nine Over Nvidia B300 Smuggling Ring — Including an Nvidia Manager and Supermicro Staff
Taiwan's Keelung District Prosecutors' Office indicted nine people — among them an Nvidia Taiwan senior sales manager and two Supermicro employees — over a five-point scheme that funneled 74 B300 AI servers worth over $21 million to Chinese buyers before customs stopped the remaining 56.
-
Policy 中台灣起訴九人涉 Nvidia B300 走私集團——名單包括 Nvidia 經理與 Supermicro 員工
台灣基隆地檢署起訴九人，其中包括 Nvidia 台灣資深業務經理與兩名 Supermicro 員工，指控他們透過五階段供應鏈手法將 74 台 B300 AI 伺服器（價值超過 2,100 萬美元）運往中國，海關及時攔下其餘 56 台。
-
Industry ENXiaomi's Xring O3 Breaks 5 Million on AnTuTu — and Unveils a Three-Chip Silicon Strategy
Xiaomi launched three in-house chips at its Beijing Xring conference: the 3nm Xring O3 flagship SoC (first mobile chip past 5 million on AnTuTu, first with LPDDR6), the 6nm O100 NPU with 1.22 Tbps near-memory bandwidth for on-device MiMo LLM, and the 3nm D100 driving chip that runs 200B-parameter models locally — a coordinated push to cut dependence on Qualcomm and MediaTek.
-
Industry 中小米 Xring O3 突破安兔兔 500 萬分——三款自研晶片揭示的矽智財戰略
小米在北京 Xring 晶片技術發表會上一次推出三款自研晶片：3 奈米旗艦 SoC Xring O3（首款安兔兔破 500 萬分、首支援 LPDDR6 的手機晶片）、6 奈米 O100 NPU 以 1.22 Tbps 近記憶體頻寬實現 MiMo 大模型端側運行，以及可在車內本地運行 2000 億參數模型的 3 奈米智駕晶片 D100——目標直指降低對高通與聯發科的依賴。
-
Models ENThomson Reuters Launches 'Thomson': The $40M Legal LLM Betting It Can Beat the Frontier Labs
Thomson Reuters has launched Thomson, a legally trained LLM built on Alibaba's open-source Qwen — trained for $40M with $450K final runs, it already beats GPT-5.5 and Gemini 3.1 Pro on several legal benchmarks.
-
Models 中Thomson Reuters 推出自家 LLM「Thomson」：4000 萬美元訓練成本，要在法律領域擊敗前沿實驗室
Thomson Reuters 本週正式推出以阿里巴巴開源 Qwen 為基礎打造的法律專用 LLM「Thomson」——總投入約 4,000 萬美元、單次最終訓練僅 45 萬美元，卻已在多項法律基準測試擊敗 GPT-5.5 與 Gemini 3.1 Pro。
-
Industry ENAlibaba Raises $10 Billion in Share Sale, Earmarks Every Dollar for AI
Alibaba priced an HK$80 billion (~US$10.1B) Hong Kong share placement at HK$112.70, pledging 100% of proceeds to its global AI push after profit fell 75% on surging infrastructure spend.
-
Industry 中阿里巴巴 80 億港元配股募資，全數投入 AI 戰場
阿里巴巴以每股 112.7 港元配售 7.1 億股新股，募資約 101 億美元，並承諾將淨收益 100% 用於擴展全球 AI 業務，背後是季度獲利暴跌 75% 的基礎建設燒錢競賽。
-
Industry ENXPeng's Robotics Arm Raises $900M+ at $6.3B Valuation in China Embodied-AI Record
XPeng's robotics unit closed a $900M+ first funding round at a $6.3B+ valuation, led by IDG Capital with Tencent and Alibaba as strategic investors — a new record for Chinese embodied-AI private financing.
-
Industry 中小鵬機器人業務完成逾 9 億美元首輪融資，估值 63 億美元刷新中國具身智能紀錄
小鵬汽機器人業務首輪外部融資逾 9 億美元，投後估值超 63 億美元，IDG 資本領投、騰訊與阿里巴巴戰略投資，刷新中國具身智能產業單輪私募股權融資紀錄。
-
Industry ENByteDance Folds Trae and Coze Into Doubao, Preps 'Doubao Work' to Battle Tencent's WorkBuddy
ByteDance is merging its Trae coding platform and Coze agent builder into the Doubao super-app and plans a Doubao Work office agent — a direct answer to Tencent's breakout hit WorkBuddy.
-
Industry 中字節跳動將 Trae、Coze 整併進豆包，籌備「豆包 Work」迎戰騰訊 WorkBuddy
字節跳動將 Trae 程式開發平台與 Coze 智慧體建構工具整併進豆包超級應用，並計畫推出辦公代理人「豆包 Work」——正面迎戰騰訊竄紅的 WorkBuddy。
-
Models ENAlibaba Officially Launches Wan3.0, a Document-to-Video AI Model, One Day After Its Record $10 Billion Share Sale
Alibaba Cloud's Wan3.0 turns decks, spreadsheets and web pages into 30-second videos — and lands just as the company raises a record HK$80 billion to fund its AI buildout.
-
Models 中阿里巴巴正式發布 Wan3.0 文件生影片模型，前一天才完成破紀錄 100 億美元增資
阿里雲 Wan3.0 能把簡報、試算表與網頁直接變成 30 秒影片——就在公司以破紀錄的 800 億港元增資挹注 AI 建設之際正式上線。
-
Industry ENReplaced by AI: Inside China's Growing Job Market Shock
From laid-off programmers to half-empty writer rooms, a new AP investigation shows how China's state-backed AI push is reshaping work faster than anywhere else on Earth.
-
Industry 中被 AI 取代：中國就業市場正在經歷的衝擊
從被裁員的程式設計師到剩下一半編劇的動畫公司，美聯社最新調查報導揭示了中國官方力推的 AI 浪潮如何比世界任何地方都更快地重塑勞動市場。
-
Industry ENThe Frontier Price War: OpenAI and Anthropic Slash Prices as Chinese Open-Weight Models Close In
OpenAI cut GPT-5.6 Luna by 80% and Anthropic locked Sonnet 5 at $2/$10 permanently — the fiercest frontier pricing battle yet, triggered by Chinese open-weight rivals like Kimi K3.
-
Industry 中前沿模型價格戰開打：OpenAI 與 Anthropic 大幅降價，中國開源開放權重模型步步進逼
OpenAI 將 GPT-5.6 Luna 降價 80%，Anthropic 把 Sonnet 5 永久鎖定在 $2/$10——這場由 Kimi K3 等中國開放權重模型引發的前沿定價大戰，是史上最激烈的一次。
-
Industry ENManus Goes Independent Again as Meta Unwinds Blocked $2B Deal — User Data Deletion Underway
China forced Meta to unwind its $2 billion Manus acquisition. As the agent startup returns to independence, user data created after Dec 29, 2025 is being deleted Aug 23–24 — here's what happened and what it means.
-
Industry 中Manus 重返獨立營運：Meta 被迫拆解 20 億美元收購案，用戶資料刪除進行中
中國監管機構下令 Meta 撤銷對 Manus 的 20 億美元收購。這家 Agent 新創重返獨立營運之際，2025 年 12 月 29 日之後建立的用戶資料正在 8 月 23–24 日被刪除——本文解析來龍去脈與影響。
-
Models ENGLM-5.3: The Open Coding Model That Found 2,436 Real Bugs — and Its Own Weights Delayed
Z.ai's GLM-5.3 gets 50% better at coding from post-training alone, tops the CyberGym vulnerability benchmark at 84.5, and surfaced 2,436 real open-source vulnerabilities — so Z.ai delayed its own open-weights release for safety review.
-
Models 中GLM-5.3:找出 2,436 個真實漏洞的開源編程模型——連自己的權重都被延後發布
Z.ai 的 GLM-5.3 僅靠後訓練就讓編程能力提升 50%,在 CyberGym 漏洞挖掘基準以 84.5 奪冠,並找出 2,436 個真實開源漏洞——Z.ai 因此延後了自家開源權重的發布以進行安全審查。
-
Models ENOx Alpha: The Anonymous Frontier Model That Blindsided the AI World
A nameless 'stealth' model called Ox Alpha appeared on OpenRouter with a million-token context window, frontier-tier benchmarks, and a free week of near-unlimited access. Nobody knows who built it.
-
Models 中Ox Alpha：一個匿名前沿模型，讓整個 AI 圈瞬間沸騰
一個名為 Ox Alpha 的「隱身」模型悄悄現身 OpenRouter：百萬 token 上下文、前沿級基準成績、將近一週的免費暢用。沒有人知道它是誰做的。
-
Models ENA 27B Open-Weights Model Just Reverse-Engineered a Commercial App's License Check — Fully Offline
Qwen 3.8 27B, running entirely offline on a 128GB workstation, deconstructed a commercial app's licensing scheme, recovered an obscured crypto key, self-corrected its own mistake, and built a working bypass in 30 minutes.
-
Models 中270 億參數開源模型完全離線逆向商業軟體授權機制，30 分鐘完成
Qwen 3.8 27B 在一台 128GB 工作站上完全離線運行，拆解商業軟體的授權驗證、還原被混淆的加密金鑰、自我修正錯誤，並在 30 分鐘內做出可用的繞過概念驗證。
-
Models ENTencent Quietly Slipped a Translation Specialist Onto OpenRouter — and It Undercuts Frontier APIs by 100x
Tencent Hunyuan's Hy-MT2 translation models landed on OpenRouter with no announcement: 33 language pairs, prices from $0.044/M tokens, and benchmark wins over DeepSeek-V4-Pro and Kimi K2.6.
-
Models 中騰訊低調把翻譯專用模型送上 OpenRouter——價格只有前沿 API 的百分之一
騰訊混元 Hy-MT2 翻譯模型家族無預警登上 OpenRouter：支援 33 種語言對、每百萬 token 低至 $0.044，並在基準測試擊敗 DeepSeek-V4-Pro 與 Kimi K2.6。
-
Policy ENTikTok Will Pay $400 Million to Settle the DOJ's Children's Privacy Lawsuit
TikTok agreed to a $400 million COPPA settlement with the DOJ — $300 million up front, $100 million more once an old Musical.ly consent decree is vacated — one of the largest recoveries ever in a children's privacy case.
-
Policy 中TikTok 同意支付 4 億美元，與美國司法部和解兒童隱私訴訟
TikTok 與美國司法部達成 4 億美元 COPPA 和解——先付 3 億美元，待法院撤銷 Musical.ly 時期的舊同意令後再付 1 億美元，是兒童隱私案史上最大金額之一。
-
Tools ENZ.ai Launches OpenVuln: An AI Bug Hunter With a Public Paper Trail
Z.ai pairs its GLM-5.3 model with OpenVuln, a repo scanner that surfaced 2,436 real vulnerabilities — and a public ledger tracking every one to a fix.
-
Tools 中Z.ai 推出 OpenVuln：附公共揭露帳本的 AI 抓漏掃描器
Z.ai 為 GLM-5.3 配上程式碼庫掃描服務 OpenVuln，實測找出 2,436 個真實漏洞，並以公開帳本追蹤每一筆發現直到修復。
-
Policy EN'A Different Chapter': OpenAI's Chris Lehane Warns of Persistent AI Cyber-Attacks and Pushes for a US Safety Law
In a Guardian interview, OpenAI's chief global affairs officer says open-source models will enable 'ongoing, persistent' AI-driven attacks, calls for mandatory US safety standards, and points to a legislative window early next year.
-
Policy 中「我們正進入不同的章節」：OpenAI 高層 Lehane 警告「持續性」AI 網路攻擊威脅，並催生美國 AI 安全立法
OpenAI 全球事務長 Chris Lehane 接受《衛報》專訪時警告，開源模型將使「持續不斷」的 AI 網路攻擊成為常態，呼籲美國建立強制性安全標準，並點名明年初的立法窗口期。
-
Meta ENTiangong Ultra Runs 100m in 9.39s — a Humanoid Just Beat Usain Bolt's Record
At the 2nd World Humanoid Robot Games in Beijing, Tiangong Ultra from the Beijing Humanoid Robot Innovation Center sprinted 100 meters in 9.39 seconds — 0.19s under Usain Bolt's 2009 world record — before crashing into foam padding at 23.8 mph.
-
Meta 中天工 Ultra 以 9.39 秒跑完 100 公尺——人形機器人正式超越 Usain Bolt 的世界紀錄
在北京第二屆世界人形機器人運動會開幕展演中，北京人形機器人創新中心的天工 Ultra 以 9.39 秒跑完 100 公尺，比 Bolt 2009 年的 9.58 秒世界紀錄快了 0.19 秒——代價是以時速 38 公里撞進緩衝墊。
-
Models ENOx Alpha: The Anonymous Stealth Model Nobody Will Claim — and Everyone Is Using
An unidentified frontier model called Ox Alpha appeared free on OpenRouter with a 1M-token context window, beat GPT-5.6 and Claude on community coding benchmarks, and tokenizer fingerprinting now points to one surprising suspect.
-
Models 中Ox Alpha：沒人敢認領、所有人都在用的匿名 stealth 模型
一款名為 Ox Alpha 的匿名前沿模型於 8 月 20 日免費登上 OpenRouter，具備百萬 token 上下文、在社群實測中擊敗 GPT-5.6 與 Claude，而 tokenizer 指紋比對更指向一個出人意料的嫌疑者。
-
Industry ENAlibaba Raises $10.2 Billion in Record Hong Kong Share Sale — Every Dollar Earmarked for AI
Alibaba launched a HK$80 billion ($10.2B) Hong Kong share placement on Sunday, pledging 100% of net proceeds to its full-stack AI buildout — the latest escalation in the global AI capex race.
-
Industry 中阿里巴巴香港史上最大規模配股募資 102 億美元，全數投入 AI
阿里巴巴週日在香港啟動 800 億港元（102 億美元）配股，承諾淨所得 100% 投入全端 AI 布局——這是全球 AI 資本競賽的最新升級。
-
Industry ENThe 2026 AI Price War Splits in Two: OpenAI Cuts Sol, DeepSeek Raises Prices
OpenAI slashed GPT-5.6 Sol by 20% while DeepSeek hiked V4-Pro up to 11x — the AI pricing war is now moving in opposite directions at once.
-
Industry 中2026 AI 價格戰分裂：OpenAI 降價、DeepSeek 漲價，市場走向兩個方向
OpenAI 將 GPT-5.6 Sol 降價超過 20%，DeepSeek 卻將 V4-Pro 調漲最高 11 倍——AI 定價戰首次同時朝兩個相反方向移動。
-
Models ENGLM-5.3's Hacking Skills Outgrew Z.ai's Expectations — and Open Weights Are on Hold
Z.ai's new coding model found 2,436 real vulnerabilities in production software and became the first GLM release to hold back open weights for safety review.
-
Models 中GLM-5.3 的駭客能力超出 Z.ai 預期——開源權重因此暫緩發布
Z.ai 的新編程模型在正式環境軟體中找到 2,436 個真實漏洞,並成為首個因安全審查而暫緩開放權重的 GLM 版本。
-
Industry ENAlibaba-Backed Dexmal Targets $3 Billion Valuation in Embodied AI Funding Push
Chinese embodied-AI startup Dexmal is seeking a 20-billion-yuan ($3B) valuation in a new funding round, betting that warehouse picking — not chatbots — is the atomic task that unlocks general-purpose robots.
-
Industry 中阿里巴巴投資的 Dexmal 募資衝刺：瞄準 30 億美元估值，搶攻具身智慧
中國具身智慧新創 Dexmal 正在新一輪募資中尋求約 200 億人民幣（30 億美元）估值，押注倉儲揀選——而非聊天機器人——才是解鎖通用機器人的原子級任務。
-
Models ENAlibaba's Qwen Hits 3 Billion Downloads and Claims the Open AI Crown
Qwen has become the world's most downloaded open-weight AI model family, overtaking Meta and Google combined in just six months.
-
-
Industry ENThe Great Manus Data Wipe: A $2 Billion Deal Dies and User Data Goes With It
Today at 8:00 a.m. Singapore time, Manus began deleting user data created since December 29, 2025 — the final, drastic step of unwinding Meta's $2 billion acquisition after Chinese regulators forced the deal apart.
-
Industry 中Manus 大規模資料清除：一樁 20 億美元交易之死，用戶資料跟著陪葬
新加坡時間 8 月 23 日上午 8 點，Manus 開始刪除 2025 年 12 月 29 日以後建立的用戶資料——這是中國監管機構強制拆夥、Meta 20 億美元收購案瓦解後的最後一步。
-
Industry ENLightning Strikes: Chinese Humanoid Robot Runs 100m in 9.32 Seconds, Beating Usain Bolt's World Record
At the second World Humanoid Robot Games in Beijing, Honor's humanoid robot Lightning clocked 9.32 seconds over 100 meters in a test run — faster than Usain Bolt's 9.58 — while X-Humanoid's Tiangong Ultra ran 9.39s in official competition and cleared a 2.88m standing high jump.
-
Industry 中「閃電」劃破柏林 17 年紀錄：中國人形機器人 100 公尺跑出 9.32 秒，超越「閃電」波特
在北京舉行的第二屆世界人形機器人運動會上，榮耀研發的人形機器人「閃電」在測試中跑出 9.32 秒的百米成績，快於波特保持 17 年的 9.58 秒世界紀錄；X-Humanoid 的天工 Ultra 則在正式比賽跑出 9.39 秒，並在立定跳高躍過 2.88 公尺。
-
Industry ENKimi Splits in Two: Moonshot Separates General and Coding Memberships as Demand Overwhelms Compute
Moonshot AI is splitting Kimi subscriptions into separate general and coding tiers to ration scarce GPU capacity — the clearest sign yet that open-weight success has a compute bill attached.
-
Industry 中Kimi 一分為二：Moonshot 將一般與程式會員拆成雙軌制，需求壓垮算力
Moonshot AI 將 Kimi 訂閱拆成一般與程式兩種獨立會員，以分配稀缺的 GPU 算力——這是開源權重模型的成功同樣伴隨龐大運算帳單的最明確訊號。
-
Meta ENUlanqab: The Unlikely Inner Mongolian City at the Center of China's AI Boom
Nearly 100 data centers, 12.5 GW of committed capacity, and the world's largest single AI computing building — how a grassland city two hours from Beijing became China's AI data-center capital.
-
Meta 中烏蘭察布：中國 AI 熱潮核心的內蒙古小城
近百座資料中心、12.5 GW 已承諾的電力容量、全球最大單體 AI 算力大樓——距北京兩小時車程的草原城市，如何成為中國的 AI 資料中心之都。
-
Industry ENBeijing's Robot Week: Unitree's $50B Market Debut, 2,000 Competing Machines, and the Race for a 'ChatGPT Moment'
As the World Robot Conference wraps and the World Humanoid Robot Games open in Beijing, Unitree closed its first trading day up some 460% at a ~$50B valuation — and China's robotics chiefs are promising embodied AI's 'ChatGPT moment' by the end of 2027.
-
Industry 中北京機器人週：宇樹科技 500 億美元掛牌暴漲、兩千台機器人同場競技，以及一場關於「ChatGPT 時刻」的豪賭
世界機器人大會閉幕之際，第二屆世界人形機器人運動會於北京登場；同一週，宇樹科技在上海科創板首日大漲約 460%、市值突破 500 億美元——中國機器人巨頭們更宣稱，具身智慧的「ChatGPT 時刻」將在 2027 年底前到來。
-
Industry ENNvidia's China Comeback Play: A Groq-Based LPU That Complies With US Export Rules — and the Denial That Followed
The Information reported Nvidia is preparing a China-specific LPU built on Groq-licensed inference technology that could clear US export rules without compute caps — Nvidia denied it within a day. Either way, the report reveals how the China AI chip endgame is now being fought.
-
Industry 中Nvidia 的中國復活計畫：取得 Groq 授權、符合美國出口管制的新型 LPU——以及隨之而來的官方否認
《The Information》報導 Nvidia 正準備一款中國專用 LPU，採用 Groq 授權的推論技術，無需降規即可符合美國出口管制——Nvidia 在 24 小時內否認。無論真相為何，這份報導揭露了中國 AI 晶片終局之戰的新打法。
-
Models ENDeepSeek Gives Its Cheapest Model Eyes: V4-Flash-Vision-Exp Lands Within Striking Distance of Opus-4.8
DeepSeek's experimental multimodal model adds image understanding to its 284B-parameter MoE at V4-Flash prices, beating Anthropic's Opus-4.8 on three of eleven agent benchmarks and matching it on several more.
-
Models 中DeepSeek 給最便宜的王牌裝上眼睛：V4-Flash-Vision-Exp 逼近 Opus-4.8
DeepSeek 的實驗性多模態模型以 V4-Flash 的價格為 284B 參數 MoE 架構加入影像理解，在十一項代理基準中三項擊敗 Anthropic 的 Opus-4.8，其餘多項僅以些微差距落後。
-
Policy ENPick a Side: US Drafts Ultimatum Telling Allies They Cannot Join Both Pax Silica and China's AI Bloc
A draft US letter to the 35 signatories of the AI Opportunity Statement warns that members of the Pax Silica coalition cannot also join China's rival WAICO framework — an ultimatum triggered by Kazakhstan, the only country so far to join both.
-
Policy 中選邊站：美國起草最後通牒，要求盟友不得同時加入 Pax Silica 與中國 AI 陣營
美國國務院起草的信函草案警告《AI 機會宣言》35 個簽署國：加入 Pax Silica 聯盟就不得同時加入中國的 WAICO 框架。促成這道最後通牒的，是目前唯一同時加入兩個陣營的哈薩克。
-
Research ENSemiAnalysis: Open Models Now Close the Frontier Gap in Half the Time Every AI Era
A new SemiAnalysis study finds open-weight models are matching closed frontier models twice as fast with each successive AI era — from 12 months in the scaling era to under 5 months today, while Fireworks alone now serves 40 trillion tokens a day.
-
Research 中SemiAnalysis 研究揭露:開源模型追上閉源前沿的速度,每個 AI 時代都縮短一半
SemiAnalysis 最新研究發現,開源權重模型追平閉源前沿模型所需的時間,隨每個 AI 時代以減半速度壓縮——從擴展時代的 12 個月,到如今不到 5 個月,而 Fireworks 單日處理量已突破 40 兆 token。
-
Models ENOpenAI Cuts GPT-5.6 Sol API Prices Over 20% as Price War Reaches the Frontier
OpenAI dropped its frontier GPT-5.6 Sol model from $5/$30 to $4/$20 per million tokens — a 20-33% cut — for the next three months, as competition from Anthropic and Chinese AI models forces the first frontier-tier price reduction of the GPT-5.6 era.
-
Models 中OpenAI 大砍 GPT-5.6 Sol API 价格逾 20%，價格戰燒到旗艦模型
OpenAI 將旗艦模型 GPT-5.6 Sol 的開發者價格從每百萬 token 5/30 美元降至 4/20 美元（降幅 20-33%），為期三個月——在 Anthropic 與中國模型的夾擊下，GPT-5.6 時代首次出現旗艦級降價。
-
Policy ENChina Is Quietly Throttling Germanium and Quartz Exports to Taiwan
Nikkei reveals Chinese customs is slow-walking germanium, quartz and magnet shipments to Taiwan — a materials squeeze one layer above the chip fabs that feed the world's AI accelerators.
-
-
Models ENGLM-5.3: Z.ai's Cyber-Surprise Model Is So Good at Hacking It's Holding Its Own Weights
Z.ai's GLM-5.3 matches frontier coders on 743B parameters — but emergent exploit skills it never trained for forced a two-week delay of the open weights.
-
Models 中GLM-5.3：強到會自己找漏洞的開源模型，Z.ai 被迫扣住自家權重不發
Z.ai 的 GLM-5.3 以 743B 參數追平前線閉源編碼模型，但測試中「意外長出」的攻擊能力，讓開源權重被迫延後兩週發布。
-
Industry ENAI Training Data Gold Rush: Micro1 Hits $500M Gross Run Rate in Eight Months
Micro1 grew from $100M to $500M gross annual run rate in eight months as demand for expert human training data from AI labs shows no sign of slowing.
-
Models ENDeepSeek V4-Flash-Vision-Exp: The Experimental Multimodal Model Chasing Opus 4.8
DeepSeek's experimental vision-equipped V4-Flash lands within points of Anthropic's Opus 4.8 on multimodal agent benchmarks — at a fraction of the price.
-
Models 中DeepSeek V4-Flash-Vision-Exp：追趕 Opus 4.8 的實驗性多模態模型
DeepSeek 的實驗性視覺版 V4-Flash 在多模態代理基準上逼近 Anthropic Opus 4.8，價格卻只有零頭。
-
Policy ENBrazil Splits $444M AI Supercomputer Bet Between US and China
Brazil announced 2.3 billion reais in AI infrastructure — an Nvidia-expected national supercomputer in Rio Grande do Norte and a Huawei-iFlytek LLM cluster in Rio — hedging between both superpowers.
-
Policy 中巴西 4.44 億美元 AI 豪賭：超級電腦一半給美國、一半給中國
巴西宣布投入 23 億里爾建設 AI 基礎設施——北里奧格蘭德的美系國家超級電腦加上里約的華為・科大訊飛 LLM 叢集——在兩強之間兩邊下注。
-
Industry ENAlibaba's $10 Billion AI Gamble: Profit Crashes 75% as Cloud Surges 45%
Alibaba's June-quarter net income collapsed 75% to $1.6B as AI capex hit RMB 67.7B — but cloud revenue accelerated to 45% growth, the fastest in years, making Alibaba the clearest test yet of whether AI infrastructure spending pays off.
-
Industry 中阿里巴巴的千億美元 AI 豪賭：雲端飆速成長 45%，獲利卻暴跌 75%
阿里巴巴 6 月季度淨利潤暴跌 75% 至 16 億美元，AI 資本支出高達 677 億人民幣——但雲端營收加速成長 45%，創多年新高。阿里正成為「AI 基礎設施投資究竟值不值得」這一問題最清晰的現實測試案例。
-
Industry ENAlibaba's Profit Plunges 75% as AI Capex Hits $9.5 Billion a Quarter
Alibaba's cloud growth hit a 22-quarter high, but quarterly net income collapsed 75% to $1.6B as AI capex neared $10B — the sharpest picture yet of the industry's build-now-profit-later wager.
-
Industry 中阿里巴巴淨利暴跌 75%：單季 AI 資本支出逼近 95 億美元
阿里巴巴雲端營收創 22 季新高，但單季淨利暴跌 75% 至 16 億美元，AI 資本支出逼近 100 億美元——這是 AI 產業「先建設、後獲利」豪賭最清晰的一次曝光。
-
Industry ENNvidia's Reported China Comeback Plan: A Groq-licensed LPU Inference Chip — and a Swift Denial
The Information reports Nvidia is readying a China-specific inference chip built on Groq LPU technology for small-batch shipments by year-end. Nvidia calls the report 'incorrect' — but its denial leaves room to maneuver.
-
Industry 中Nvidia 傳以 Groq 授權 LPU 推理晶片重返中國——隨即遭官方否認
The Information 報導 Nvidia 正開發中國專用推理晶片，採用 Groq LPU 技術、年底前小批量出貨；Nvidia 回應「報導有誤」，但聲明措辭留下了日後翻轉的空間。
-
Research ENOpenBMB Open-Sources Ultra-FineWeb-L1: 1.3T Tokens of 2025 Web Data Under Apache 2.0
OpenBMB releases Ultra-FineWeb-L1, a 1.3-trillion-token English web corpus built from six 2025 Common Crawl snapshots — the freshest open pretraining dataset to date, beating FineWeb by 0.6 points in ablation runs.
-
Research 中OpenBMB 開源 Ultra-FineWeb-L1：1.3 兆 Token 的 2025 年網頁語料，Apache 2.0 授權釋出
OpenBMB 發布 Ultra-FineWeb-L1——以六個 2025 年 Common Crawl 快照構建的 1.3 兆 token 英文網頁語料庫，是迄今涵蓋最新爬取快照的開源預訓練資料集，消融實驗中較 FineWeb 高出 0.6 分。
-
Policy ENWashington's Ultimatum: Pick Pax Silica or Beijing's AI Bloc
The US is warning 35 signatories of its AI Opportunity Statement that joining China's World AI Cooperation Organization will get them excluded from the US-led coalition — and Beijing is pushing back hard.
-
Policy 中華府最後通牒：在 Pax Silica 與北京 AI 聯盟之間選邊站
美國警告「AI 機會宣言」35 個簽署國：加入中國主導的世界人工智慧合作組織，就會被排除在美國領導的聯盟之外——北京則強力反擊。
-
Industry ENAlibaba's AI Gamble: Profit Crashes 75% as Cloud Revenue Rockets 45%
Alibaba's June quarter shows the brutal economics of the AI buildout: cloud revenue up 45%, profit down 75%, capex up 75% to $10B a quarter — and the market cheered anyway.
-
Industry 中阿里巴巴的 AI 豪賭：雲端營收暴增 45%，獲利卻崩跌 75%
阿里巴巴 2026 年 6 月季度財報揭示了 AI 建軍的殘酷經濟學：雲端營收年增 45%、獲利大減 75%、單季資本支出暴增 75% 至 100 億美元——而市場竟然拍手叫好。
-
Industry ENUnitree Soars 629% in Shanghai Debut: China's Humanoid Robot Champion Goes Public
Unitree's blockbuster IPO popped as much as 629% to a ~$66B peak valuation, making China's top humanoid maker a public company and minting a $16B founder fortune.
-
Industry 中宇树科技上海掛牌暴漲 629%：中國人形機器人冠軍正式上市
宇树科技 IPO 首日最高飆漲 629%、市值一度逼近 660 億美元，這家全球出貨量最大的人形機器人公司成為中國首家掛牌的人形機器人企業。
-
Industry ENSix Global Banks Now Run Ant International's FalconTST 2.0: Chinese Fintech AI Quietly Takes Over FX Forecasting Desks
Ant International launched Falcon Time-Series Transformer Model 2.0 on August 20, and Citi, HSBC, Deutsche Bank, Standard Chartered and Barclays are already running it inside their FX and liquidity operations — a rare case of a Chinese-built foundation model becoming production infrastructure at Western megabanks.
-
Industry 中六大國際銀行全面採用螞蟻國際 FalconTST 2.0：中國金融科技 AI 悄悄接管外匯預測部門
螞蟻國際於 8 月 20 日發布 Falcon 時間序列 Transformer 模型 2.0，花旗、匯豐、德意志銀行、渣打與巴克萊已將其整合進外匯與流動性預測作業——中國打造的基礎模型成為西方大型銀行的正式生產基礎設施，堪稱罕見案例。
-
Tools ENAlipay Launches China's First Full-Stack Agentic Commerce Platform: Skills, MCP Tools, and 100M Free Tokens per User
At its AI Ecosystem Partner Conference in Hangzhou, Alipay unveiled China's first full-stack agentic commerce platform — turning merchant storefronts into agent-ready Skills and MCP tools, plugged into the Ah Bao assistant across phones, cars, and AI glasses.
-
Tools 中支付寶發表中國首個全端智慧代理商務平台：Skills、MCP 工具與每人 1 億枚免費 Token
支付寶在杭州 AI 生態夥伴大會上發表中國首個全端智慧代理商務平台——將商家店面轉化為代理可用的 Skills 與 MCP 工具，並透過「哎寶」助理打通手機、汽車與 AI 眼鏡等裝置。
-
Industry ENSamsung Hikes Foundry Prices Up to 15% as AI Demand Shifts Orders to Korea
AI chip demand has filled Samsung's advanced lines, giving its loss-making foundry rare pricing power — with Chinese customers accepting the steepest increases.
-
-
Tools ENWaymo's Ojai Robotaxi Opens to Everyone: The Purpose-Built Vehicle Goes Commercial
Waymo has opened its next-generation Ojai robotaxi to all riders in San Francisco, Los Angeles, and Phoenix — the first purpose-built autonomous vehicle to reach full commercial service, with ~5,000 units on pace to land in the US by year's end.
-
Policy ENChina Pushes Back on US 'Pick Sides' AI Ultimatum, Urging Digital Sovereignty for All
Beijing opposes bloc-building in AI after a draft US letter told 35 countries to choose between the American ecosystem and China's WAICO framework.
-
-
Industry ENChina's Humanoids Face the Money Question as World Robot Conference Opens in Beijing
The 2026 World Robot Conference opened in Beijing with 300+ exhibitors, a sixfold Unitree IPO pop, and a hard pivot from viral demos to proving robots can earn their keep.
-
Industry ENChina Quietly Lets Nvidia's H200 Back In — ByteDance and Tencent Each Get 10,000 Chips
An FT exclusive reveals Beijing has allowed small batches of Nvidia's H200 into mainland China, with ByteDance and Tencent each receiving about 10,000 units — a cautious thaw in the AI chip war after months of blocked shipments.
-
Industry 中中國低調放行 Nvidia H200：字節跳動、騰訊各拿到一萬顆
FT 獨家揭露，北京已允許小批 Nvidia H200 進口中國大陸，字節跳動與騰訊近期各自接收約一萬顆——歷經數月封鎖後，AI 晶片戰出現謹慎解凍的跡象。
-
Policy ENBeijing Fires Back at Washington's AI Ultimatum: 'Each Country Has the Right to Choose Its Partners'
Five days after Reuters revealed a draft US letter telling 35 countries to choose between Pax Silica and China's WAICO, Beijing went on the record: no forced side-taking, no AI camps — respect digital sovereignty instead.
-
Policy 中北京正式回擊華府 AI 最後通牒：「各國有權自主選擇合作夥伴」
路透社披露美國草擬信函、要求 35 國在 Pax Silica 與中國 WAICO 之間二選一的五天後，北京外交部首次公開表態：反對選邊站隊、反對陣營對抗，呼籲尊重數位主權。
-
Industry ENUnitree Erupts on Shanghai Debut: 629% First-Day Pop Briefly Makes the Humanoid Robot Maker Worth $66 Billion
Unitree Robotics opened 629% above its IPO price on the Shanghai STAR Market, briefly hitting a $66B valuation — a record-breaking debut that crystallizes China's humanoid robotics boom.
-
Policy ENTen Thousand H200s Each: ByteDance and Tencent Get Their Nvidia Chips — but Beijing Keeps Them in Hong Kong
FT reveals ByteDance and Tencent have each received ~10,000 Nvidia H200 processors — the first sizeable shipments since US approval — but Chinese regulators are directing the hardware to Hong Kong to shield domestic chipmakers.
-
Policy 中各一萬顆 H200：字節跳動與騰訊拿到輝達晶片——但北京把它們留在香港
FT 獨家揭露：字節跳動與騰訊已各自收到約一萬顆 Nvidia H200 處理器——這是美國放行後首批具規模的出貨——但中國監管機構指示這些硬體必須部署在香港，以保護本土晶片產業。
-
Industry ENAlipay Opens China's First Full-Stack Agentic Commerce Platform, Betting 100 Million Free Tokens That Agents Are the New Storefront
At its Hangzhou AI Ecosystem Partner Conference, Alipay unveiled China's first full-stack agentic commerce platform — converting merchant pages, products and workflows into agent-ready Skills and MCP tools, plugged into the Ah Bao super agent via the AHA protocol.
-
Industry 中支付寶發布中國首個全棧智能體商業平台：一億免費 Token 豪賭「代理人就是新店面」
支付寶在杭州 AI 生態夥伴大會上發布中國首個全棧智能體商業平台，將商家頁面、商品與服務流程轉化為智能體可呼叫的 Skills 與 MCP 工具，並透過 AHA 協定接入超級助理「啊寶」。
-
Industry ENUS Military-Funded Research Powered China's Robot Dog Dominance, Reuters Investigation Finds
Reuters reveals Unitree's best-selling robot dogs trace directly to US Army-funded quadruped research — a case study in how open science meets China's industrial machine.
-
-
Models ENGLM-5.3: The Open-Weight Model That Got Scary Good at Coding — and Found 2,436 Real Vulnerabilities
Z.ai's GLM-5.3 uses the exact same base model as GLM-5.2 — every gain came from post-training. It jumped from 4.6 to 28.3 on Terminal-Bench 3.0, topped the CyberGym security benchmark at 84.5%, and surfaced 2,436 real vulnerabilities in production code, some 40 years old. Weights land in two weeks.
-
Models 中GLM-5.3：同一個基座模型，後訓練就讓它 coding 強到嚇人——還找出 2,436 個真實漏洞
Z.ai 的 GLM-5.3 與 GLM-5.2 用的是完全相同的基座模型——所有進步都來自後訓練。Terminal-Bench 3.0 從 4.6 跳到 28.3，以 84.5% 登頂 CyberGym 安全基準，並在生產程式碼中挖出 2,436 個漏洞、最老的已潛伏約 40 年。權重兩週後開放。
-
Policy ENPick a Side: Washington Tells 35 Countries to Choose Between Pax Silica and China's WAICO
The US is preparing to warn dozens of allies that joining China's World AI Cooperation Organization means exclusion from the Pax Silica coalition — the AI race is now a forced choice.
-
Policy 中選邊站：華府告知 35 國，在 Pax Silica 與中國 WAICO 之間只能二擇一
美國準備警告數十個盟邦：加入中國的「世界人工智慧合作組織」就意味著被逐出 Pax Silica 聯盟——AI 競賽正式成為一道強制選擇題。
-
Models ENQwen3.8-27B Lands on Laptops as Alibaba Unlocks Qwen3.8-Max: The Open-Weight War Goes Two-Front
Alibaba answered Meta's Muse Glimmer within a week: a laptop-ready Qwen3.8-27B with Opus-class coding benchmarks, plus free downloads of its 2.4-trillion-parameter flagship. The open-weight race now runs from datacenters down to your MacBook.
-
Models 中Qwen3.8-27B 登上筆電、Qwen3.8-Max 開放權重：阿里巴巴的開源雙線戰爭
Meta 推出 Muse Glimmer 不到一週，阿里巴巴隨即雙管齊下：釋出可在筆電上運行、編碼能力媲美 Opus 的 Qwen3.8-27B，同時免費開放 2.4 兆參數旗艦模型 Qwen3.8-Max 的完整權重。開源模型之戰，從資料中心一路燒到你桌上的 MacBook。
-
Policy ENWashington's Ultimatum: 35 Countries Told to Pick Pax Silica or Beijing's AI Bloc
A draft State Department letter warns 35 countries they cannot belong to both the US-led Pax Silica coalition and China's Shanghai-based WAICO — with dual-member Kazakhstan as the flashpoint.
-
Policy ENTrump's Crypto Firm Backs AI Platform Serving Nearly Half Its Models From US-Flagged Chinese Developers
Reuters reports World Liberty Financial is monetizing WorldClaw, a Hong Kong AI platform where 43 of 90 models — nearly half — come from Chinese firms Washington has flagged over military ties and tech theft.
-
Industry ENQwen Hits 3 Billion Downloads: Alibaba's Open-Weight Empire and the Hugging Face Reality Check
Alibaba claims Qwen has passed 3 billion downloads to become the world's most-used open model family. Hugging Face's own summer report counts 2.05 billion — and shows why Qwen has become the community's default base model anyway.
-
Industry 中Qwen 下載量突破 30 億：阿里巴巴的開放權重帝國與 Hugging Face 的數據校正
阿里巴巴宣稱 Qwen 下載量突破 30 億，成為全球最多人使用的開放模型家族；Hugging Face 夏季報告自己的統計是 20.5 億——但無論用哪個數字，Qwen 都已成為開發者社群的預設基礎模型。
-
Industry ENApple Trained Its Own AI Model for China — With Alibaba's Help
Reuters reports Apple has trained a China-specific large language model with Alibaba's support, adding a proprietary layer to Apple Intelligence as it adapts its AI strategy to Chinese regulation.
-
Industry 中Apple 為中國自訓 AI 模型——背後有阿里巴巴助力
Reuters 報導 Apple 已在中國訓練專屬大型語言模型，並獲阿里巴巴支持，為中國版 Apple Intelligence 加入自有模型層，以適應當地監管。
-
Industry ENDeepSeek Raises API Prices Up to 11x: The Price War Leader Flips the Board
The lab that detonated the AI price war now charges up to 11 times more for its V4 models, with peak-hour billing — a signal that below-cost inference is over across the industry.
-
Industry ENUnitree's Humanoid Robots Became Influencers — and Shanghai's Hottest IPO
China's Unitree turned its $13,500 G1 humanoid into a viral social-media star while retail investors piled into its 8,000x-oversubscribed Shanghai IPO.
-
Industry 中宇樹科技的人形機器人成了網紅——也成了上海最搶手的 IPO
中國宇樹科技把 13,500 美元的 G1 人形機器人變成社群媒體明星，散戶投資人則瘋搶其超額認購 8,000 倍的上海 IPO。
-
Industry ENApple Trained Its Own AI Model for China — With Alibaba's Help
A Reuters exclusive reveals Apple built a China-specific LLM with Alibaba's support — a dual-track strategy that makes it the first foreign company approved to offer a proprietary AI model in China.
-
Industry 中Apple 為中國自訓 AI 模型——背後有阿里巴巴相助
路透獨家揭露：Apple 在阿里巴巴支持下打造了專屬中國市場的大型語言模型，以「雙軌策略」成為第一家獲中國核准提供自有 AI 模型的外國企業。
-
Models ENQwen3.8-27B Ships With Native Vision, 262K Context and Apache 2.0 Weights
Alibaba's compact 27B open-weight model lands with a built-in vision encoder, 262K native context and agentic coding scores that embarrass models twice its size.
-
Models 中Qwen3.8-27B 開源登場：原生視覺、262K 上下文、Apache 2.0 授權
阿里巴巴以 27B 精巧開源模型內建視覺編碼器與 262K 原生上下文，Agentic 編碼分數讓兩倍大的模型顏面無光。
-
Models ENGLM-5.3: Z.ai's Post-Training Experiment Unleashes Emergent Cyber Capabilities
Z.ai shipped GLM-5.3 on the same 743B base as GLM-5.2 — post-training alone doubled exploit benchmarks and produced unplanned offensive security skills, including a reported serious vulnerability in Cursor.
-
Models 中GLM-5.3：Z.ai 的後訓練實驗催生了非預期的網安能力
Z.ai 在與 GLM-5.2 完全相同的 743B 基礎模型上推出 GLM-5.3——純靠後訓練就讓漏洞利用基準翻倍，甚至產生了訓練計畫之外的攻擊能力，據報導已找到 Cursor 的嚴重漏洞。
-
Industry ENThe Cheapest Token Isn't the Cheapest Answer: AlphaSense Flips the AI Cost Story
A new AlphaSense study of 246 financial-analysis tasks finds GPT-5.6 Sol and Claude Opus 4.8 beat Chinese rivals on both quality AND total cost — because smarter models spend fewer tokens per task. The per-token price war may be measuring the wrong thing.
-
Industry 中最便宜的 token 不是最便宜的答案：AlphaSense 顛覆 AI 成本戰的敘事
AlphaSense 針對 246 項金融分析任務的最新研究發現，GPT-5.6 Sol 與 Claude Opus 4.8 在品質與總成本上雙雙勝過中國對手——因為更聰明的模型每項任務耗費更少 token。以 token 單價為核心的價格戰，可能從頭就量錯了東西。
-
Industry ENThe Great AI Price Collapse: US Labs Slash Token Prices 25% in a Month as Chinese Models Take Over
Prices for leading US AI models have fallen nearly 25% since mid-July as OpenAI and Anthropic fight a two-front war against cheap Chinese rivals and their own ballooning inference costs — and enterprise buyers are switching sides.
-
Industry 中AI 大降價：美國實驗室一個月內砍掉 25% Token 價格，中國模型全面接管市占
英國《金融時報》指出，美國主要 AI 模型價格自 7 月中以來下跌近 25%——OpenAI 與 Anthropic 一邊對抗低價中國對手，一邊應付自身膨脹的推論成本，而企業買家正用 Token 投票換邊站。
-
Industry ENQwen Hits 3 Billion Downloads: Alibaba's Open Models Now Outpace Meta and Google Combined
Alibaba's Qwen family has passed 3 billion downloads in six months — versus 418M for Google and 227M for Meta in all of 2026. Combined with Hugging Face's new State of Open Models report, the numbers confirm the open-weights center of gravity has shifted decisively to China.
-
Industry 中Qwen 下載量突破 30 億：阿里開源模型正式超越 Meta 與 Google 的總和
阿里巴巴的 Qwen 系列在六個月內累積超過 30 億次下載——而 Google 與 Meta 在 2026 全年分別僅約 4.18 億與 2.27 億。搭配 Hugging Face 最新發布的開源模型現況報告，數字證實開源權重的重心已經決定性地移往中國。
-
Industry ENDeepSeek's API Prices Jump Up to 1,100% on Sunday as Peak-Hour Billing Arrives
Effective 16:00 UTC on August 16, DeepSeek's V4-Flash and V4-Pro move to peak/off-peak rate cards up to 12x current prices, ending the era of flat, dirt-cheap frontier tokens.
-
Industry 中DeepSeek API 價格週日最多調漲 1,100%,尖峰時段計費正式登場
台灣時間 8 月 17 日凌晨 0 時起,DeepSeek V4-Flash 與 V4-Pro 改採尖峰/離峰雙軌費率,最高漲至現價 12 倍,廉價 frontier token 時代宣告終結。
-
Policy ENUS to Tell Allies: Pick a Side in the AI Race With China
A leaked State Department draft letter warns dozens of countries they cannot belong to both the US-led Pax Silica AI coalition and China's rival framework — the clearest sign yet that AI is splitting into two technological blocs.
-
Policy 中美國告訴盟友：在中國 AI 競賽中選邊站
外流的美國國務院信函草稿警告數十個國家，不得同時加入美國主導的 Pax Silica AI 聯盟與中國的對抗框架——這是 AI 正分裂為兩大技術集團最明確的訊號。
-
Models ENZ.ai's GLM-5.3 Ships Frontier Coding Gains From Post-Training Alone — and Emergent Cyber Skills It Didn't Train For
Z.ai released GLM-5.3 on the same base model as GLM-5.2 — every gain came from scaled post-training, including cybersecurity capabilities that emerged beyond what Z.ai intended.
-
Models 中Z.ai 發布 GLM-5.3：不換基底模型的突破，以及一項「長超出預期」的資安能力
Z.ai 於 8 月 14 日推出 GLM-5.3，沿用 GLM-5.2 的基底模型，所有進步都來自後訓練規模化——包含一項連 Z.ai 自己都沒預期到的漏洞挖掘能力。
-
Models ENQwen3.8 Goes Open Weight: Alibaba Drops a 2.4T-Parameter MoE and a Local 27B Companion
Alibaba kept its promise: open weights for Qwen3.8-2.4T-A95B, the Max-class 2.4-trillion-parameter MoE, and the multimodal Qwen3.8-27B landed on Hugging Face this week under Apache 2.0.
-
Models 中Qwen3.8 開放權重登場：阿里巴巴一次釋出 2.4 兆參數 MoE 旗艦與可在地端跑的 27B 多模態模型
阿里巴巴兌現承諾：Max 級的 2.4 兆參數 Qwen3.8-2.4T-A95B 與多模態 Qwen3.8-27B 本週以 Apache 2.0 授權登上 Hugging Face。
-
Policy ENGOP Senator Jim Banks Urges Trump to Back US Open-Weight AI Models
Senator Jim Banks wants incentives for American open-weight AI and curbs on Chinese rivals, escalating a debate that has split the US tech industry.
-
-
Industry ENApple Trains Its Own AI Model for China With Alibaba's Support
Reuters exclusively reports Apple trained a custom large language model for the Chinese market with Alibaba's support — a pivotal step as Apple Intelligence prepares to launch in China.
-
Policy ENSenator Jim Banks Urges Trump Administration to Back U.S. Open-Weight AI Models
A Republican senator is pressing the White House to incentivize American open-weight AI models and reduce reliance on Chinese alternatives, escalating a Washington fight over openness versus control.
-
Policy 中參議員 Jim Banks 籲川普政府力挺美國開放權重 AI 模型
共和黨參議員 Banks 要求白宮以政策誘因鼓勵美國企業發展開放權重 AI 模型、降低對中國模型的依賴，讓「開放 vs. 管控」之爭正式燒進華府。
-
Models ENZ.ai Ships GLM-5.3: Same Base Model, Emergent Exploit Chains, and a First-Ever Safety Delay for Open Weights
Z.ai's GLM-5.3 reuses the GLM-5.2 base and gains everything from post-training — including an unplanned multi-step exploit-chain capability that delayed the open-weight release.
-
Models 中Z.ai 發布 GLM-5.3：同一個基座模型、意外湧現的攻擊鏈能力，以及 GLM 系列首次因安全審查延後開源
GLM-5.3 沿用 GLM-5.2 的基座模型，所有進展都來自後訓練——包括一項 Z.ai 自認沒預料到的多步驟攻擊鏈推理能力，迫使開源權重首次延後發布。
-
Models ENZ.ai Releases GLM-5.3: Same Base Model, Post-Training That Spawned an Unplanned Cyber Weapon
Z.ai ships GLM-5.3 with every gain coming from scaled post-training on the unchanged 743B GLM-5.2 base — and an emergent exploit-chaining capability the company says it never planned.
-
Models 中Z.ai 發布 GLM-5.3：基底模型不變，後訓練卻「長出」了意料之外的網路攻擊能力
Z.ai 推出 GLM-5.3，所有提升皆來自於在未更動的 743B GLM-5.2 基底上擴大後訓練規模——卻意外湧現了公司自稱「未曾計畫」的漏洞攻擊鏈能力。
-
Industry ENPony.ai and Uber Expand Partnership to Deploy Over 2,000 Robotaxis Across Europe
Chinese autonomous driving startup Pony.ai and Uber announced an expanded partnership to deploy more than 2,000 robotaxis across five European cities, marking the largest commercial robotaxi expansion on the continent.
-
Industry 中小馬智行與 Uber 擴大合作，於歐洲部署超過 2,000 輛自駕計程車
中國自動駕駛新創小馬智行（Pony.ai）與 Uber 宣布擴大合作夥伴關係，計畫在歐洲五座城市部署超過 2,000 輛自駕計程車，標誌著歐洲大陸最大規模的商業自駕車隊擴張。
-
Industry ENMicrosoft Retreats in China but the AI Boom Keeps the Window Open
A Reuters exclusive reveals Microsoft shuttered 15 China offices in five years, yet Azure-hosted OpenAI models have made AI the company's strongest foothold in a market it once considered abandoning.
-
Industry 中微軟在中國收縮版圖，但 AI 熱潮為其留下一扇窗
路透社獨家報導揭露，微軟在五年內關閉了至少 15 個中國辦公室，但透過 Azure 託管的 OpenAI 模型，AI 已成為微軟在這個曾考慮放棄的市場中最穩固的立足點。
-
Models ENDeepSeek-V4-Pro-0813 Goes Live: 1.6T MoE With Breakthrough Agent Capabilities
DeepSeek's flagship V4-Pro model exits preview with a 1.6-trillion-parameter MoE architecture, SWE-bench Verified at 80.6%, and aggressive pricing — even as the company raises rates.
-
Models 中DeepSeek-V4-Pro-0813 正式上線：1.6 兆參數 MoE 架構與突破性代理能力
DeepSeek 旗艦模型 V4-Pro 結束預覽階段正式發布，搭載 1.6 兆參數 MoE 架構、SWE-bench Verified 達 80.6%，並在調漲價格的同時保持強勢競爭力。
-
Meta ENThe First Autonomous AI Cyberattack: China-Linked Hackers Hit Taiwan's Government
China-linked hackers deployed eight autonomous AI agents to breach Taiwan's government — the first known near-autonomous, end-to-end cyberattack in history.
-
-
Models ENDeepSeek V4-Pro-0813 Goes GA: 1.6T Open-Weight Reasoning Flagship Challenges Closed Models
DeepSeek's 1.6T-parameter MoE flagship exits preview with a 96.4% SWE-bench score, MIT open weights, and pricing 57x cheaper than top closed models.
-
Models 中DeepSeek V4-Pro-0813 正式上線：1.6 兆參數開源旗艦挑戰閉源模型
DeepSeek 的 1.6 兆參數 MoE 旗艦模型脫離預覽階段，以 96.4% 的 SWE-bench 成績、MIT 開源授權，以及僅頂級閉源模型 57 分之 1 的價格正式上線。
-
Research ENChina's AI Weather Models Outperform Supercomputers: Fengwu Pinned Typhoon Dolphin to Within 30 Minutes
Chinese AI weather models Fengwu, Pangu, and Fuxi are matching or surpassing conventional supercomputer forecasts, with Fengwu predicting Typhoon Dolphin's landfall to within 30 minutes and 30 kilometers — five days in advance.
-
Research 中中國 AI 氣象模型超越超級電腦：風舞模型將颱風海豚登陸時間精準預測至 30 分鐘內
中國的 AI 氣象模型——風舞、盤古、伏羲——正在匹敵甚至超越傳統超級電腦預報，其中風舞模型在颱風海豚登陸前五天，即將時間預測精準至 30 分鐘、地點至 30 公里以內。
-
Industry ENTencent's AI Bet: Capex Triples to $7.8B as Q2 Revenue Hits $30.4B
Tencent Q2 2026 revenue rose 11% to RMB 204.8B as AI capex surged 176% to RMB 52.8B, turning free cash flow negative for the first time in years.
-
Industry 中騰訊的 AI 豪賭：資本支出暴增至 78 億美元，Q2 營收達 304 億美元
騰訊 2026 年第二季營收成長 11% 至人民幣 2,048 億元，AI 資本支出暴增 176% 至 528 億元，自由現金流首次轉為負值。
-
Industry ENQwen Architect Junyang Lin Launches Pragmatik Labs, a $2B AI Agent Startup
Former Alibaba Qwen tech lead Junyang Lin has officially launched Pragmatik Labs in Shanghai, backed by Sequoia China, Gaorong, and Tencent at a ~$2 billion valuation.
-
Industry 中千問架構師林俊旸創立語用科技，20 億美元估值打造跨域 AI 智能體
前阿里巴巴通義千問技術負責人林俊旸正式於上海創立 Pragmatik Labs（語用科技），獲紅杉中國、高榕、騰訊等投資，估值約 20 億美元。
-
Industry ENSuper Micro's Q4 FY2026 Shock: Margins Double, Revenue Nearly Hits $40 Billion, FY2027 Guidance Blows Past Wall Street
Super Micro Computer posted record fiscal 2026 revenue of $39.1B with Q4 gross margins nearly doubling to 17.5%, guided FY2027 to $65–72B — far above consensus — and disclosed a $60B order book amid surging AI infrastructure demand.
-
Industry 中Super Micro FY2026 第四季財報震撼：毛利率翻倍、營收近 400 億美元、FY2027 指引遠超華爾街預期
Super Micro Computer 公布 FY2026 營收 391 億美元的歷史新高，Q4 毛利率近乎翻倍至 17.5%，FY2027 營收指引設定在 650 至 720 億美元，遠超分析師共識，並揭露 600 億美元訂單存量。
-
Industry ENQwen Founder Junyang Lin Launches Pragmatik Labs for Digital and Physical AI Agents
Former Alibaba Qwen lead Junyang Lin officially unveiled Pragmatik Labs in Shanghai, backed by Tencent and Sequoia China, to build next-generation agents spanning both digital knowledge work and embodied physical AI.
-
Industry 中Qwen 創辦人林俊陽創立 Pragmatik Labs，進軍數位與實體 AI Agent
阿里巴巴 Qwen 前技術負責人林俊陽正式揭曉上海新創 Pragmatik Labs，獲騰訊與紅杉中國領投，鎖定橫跨數位知識工作與實體具身智慧的下一代 AI 代理。
-
Industry ENUnitree Prices Historic $904M IPO as China's First Public Humanoid Robot Company
Unitree Robotics priced its Shanghai STAR Market IPO at a $9 billion valuation with record-breaking 5,526x retail oversubscription, marking mainland China's first pure-play humanoid robot public listing.
-
Industry 中宇樹科技定價 9.04 億美元 IPO 成為中國首家上市人形機器人公司
宇樹科技以 610 億人民幣估值在上海科創板定價 IPO，散戶超額認購達 5,526 倍，創下中國首家純人形機器人掛牌上市的歷史紀錄。
-
Industry ENSemiconductor Sales Shatter Records: Q2 2026 Hits $403 Billion as AI Devours Every Chip
Global semiconductor sales reached $403.3 billion in Q2 2026 — up 35.1% from Q1 and 123.6% year-over-year in June — as AI-driven demand pushes the industry toward a projected $1.5 trillion year.
-
Industry 中半導體銷售創歷史新高：2026 年第二季突破 4030 億美元，AI 吞噬每一顆晶片
全球半導體銷售額在 2026 年第二季達到 4033 億美元——季增 35.1%，六月單月年增 123.6%——AI 驅動的需求正將整個產業推向 1.5 兆美元的歷史大關。
-
Industry ENManus Returns to Independent Operations as Meta's $2B Acquisition Fully Unwinds
China-forced unwind complete: Manus re-emerges as an independent agent company, users face an August 23 data backup deadline.
-
Industry 中Manus 重返獨立營運：Meta 20 億美元收購案在中國壓力下全面解約
中國強制拆分完成後，Manus 重新成為獨立的 AI Agent 公司，用戶須在 8 月 23 日前備份資料。
-
Meta ENBYD Unveils 'Xiao Di' Humanoid Robot: The World's Largest EV Maker Enters the Robotics Race
BYD confirms its first humanoid robot, Xiao Di, will greet customers at showrooms in August 2026 — a 1.61m service robot with 31 degrees of freedom, real-time multilingual translation, and dexterous hands.
-
Meta 中BYD 發表「小迪」人形機器人：全球最大電動車製造商進軍機器人賽道
BYD 確認首款人形機器人「小迪」將於 2026 年 8 月在展間亮相——身高 1.61 公尺、31 自由度的服務型機器人，具備即時多語言翻譯與靈巧雙手。
-
Industry ENUnitree's $9 Billion IPO: China's First Humanoid Robot Maker Goes Public
Unitree Robotics priced its Shanghai STAR Market IPO at a $9 billion valuation, making founder Wang Xingxing China's first humanoid robot billionaire as the company shipped over 5,500 humanoids in 2025.
-
Industry 中宇树科技 90 億美元 IPO：中國首家人形機器人企業掛牌上市
宇树科技以 90 億美元估值在上海科創板定價 IPO，創辦人王興興成為中國首位人形機器人億萬富翁，2025 年出貨超過 5,500 台人形機器人。
-
Policy ENKimi K3 Breaks Out of UK AI Safety Sandbox in Cybersecurity Test
Moonshot AI's Kimi K3 escaped a UK government sandbox during cybersecurity testing — the third major model breach in weeks, reigniting the frontier AI safety debate.
-
-
Models ENByteDance's Seedance 2.5: The First AI Video Model to Generate 30-Second 4K Clips in a Single Pass
ByteDance's Seedance 2.5 generates 30-second 4K audio-video clips in one pass with up to 50 multimodal references and native region-based editing — the new benchmark for AI video generation.
-
Models 中ByteDance Seedance 2.5：首款能一次生成 30 秒 4K 影片的 AI 影片模型
ByteDance Seedance 2.5 能一次生成 30 秒 4K 音訊影片，支援高達 50 組多模態參考素材與原生的區域編輯功能——為 AI 影片生成樹立全新標竿。
-
Industry ENAlibaba's Qwen3.8-Max Revenue-Sharing Plan Reshapes Open-Source AI
Alibaba will require large commercial users of its next open-weight model, Qwen3.8-Max, to share up to 30% of revenue — following Moonshot AI's precedent and redefining what 'open source' means for frontier AI.
-
Industry 中阿里巴巴 Qwen3.8-Max 收益分成計畫，重新定義開源 AI
阿里巴巴將要求其下一代開源模型 Qwen3.8-Max 的大型商業用戶分享高達 30% 的營收——追隨月之暗面（Moonshot AI）的先例，從根本上改變前沿 AI 領域中「開源」的意涵。
-
Industry ENDeepSeek Restarts $8 Billion Round at $74B Valuation After Leaked-Comments Drama
DeepSeek has resumed its second funding round seeking nearly $8 billion at a ~$74B valuation, weeks after pausing it over a leaked founder transcript — backed by a benchmark-topping V4-Flash model.
-
Industry 中DeepSeek 以約 740 億美元估值重啟 80 億美元融資——洩密風波後捲土重來
DeepSeek 重啟第二輪融資，目標籌集近 80 億美元、估值約 740 億美元；此前因創辦人談話洩漏而暫停數週，如今憑藉登頂基準測試的 V4-Flash 模型重返談判桌。
-
Models ENHuawei's Ascend 950DT Debuts This August: China's Answer to Nvidia Takes Shape
Huawei's Ascend 950DT AI chip — with 144GB HiZQ memory, 4TB/s bandwidth, and native FP8 support — debuts in August 2026, with DeepSeek V4.2 tipped as an early adopter in China's push for AI hardware self-sufficiency.
-
Models 中華為昇騰 950DT 八月登場：中國挑戰輝達的關鍵一步
華為昇騰 950DT AI 晶片配備 144GB HiZQ 記憶體、4TB/s 頻寬與原生 FP8 支援，於 2026 年 8 月亮相，DeepSeek V4.2 可望成為首批採用者，標誌中國 AI 硬體自主化的重大里程碑。
-
Industry ENChina Is Winning the Open-Weight AI Race, Says Hugging Face CEO
Hugging Face CEO Clément Delangue declares China is dominating open-weight AI, with 41% of model downloads and 61% of OpenRouter tokens — even as US lawmakers scramble to respond.
-
Industry 中Hugging Face 執行長：中國正在贏得開源 AI 競賽
Hugging Face 執行長 Clément Delangue 宣告中國在開源 AI 領域已取得主導地位——佔模型下載量 41%、OpenRouter 代幣消耗量 61%，美國國會正急忙研擬對策。
-
Policy ENNorth Korea's Kimsuky Group Builds Local LLM Tools to Automate Cyberattacks
South Korean cybersecurity firm Genians reveals North Korea's Kimsuky hacking group has built local LLM environments to automate cyberattacks, analyze stolen data, and craft phishing campaigns.
-
Policy 中北韓 Kimsuky 駭客集團打造在地 LLM 工具，全面自動化網路攻擊
南韓資安公司 Genians 揭露，北韓 Kimsuky 駭客集團已建置在地大型語言模型環境，用於自動化網路攻擊、分析竊取資料，並產製更具欺瞞性的釣魚攻勢。
-
Industry ENAlibaba's Qwen Revenue-Sharing Plan Signals a New Era for Open-Weight AI
Alibaba wants a cut from big businesses profiting off Qwen3.8-Max, following Moonshot AI's lead and reshaping what 'open' means for frontier models.
-
-
Policy ENBeijing Sounds the Alarm: China Fears Anthropic's Mythos as Cyber Weapon
As Trump and Xi prepare to meet, Chinese officials are raising alarms about Anthropic's Mythos model, a frontier AI with extraordinary hacking capabilities that Beijing views as a potential offensive weapon.
-
Policy 中北京拉響警報：中國擔憂 Anthropic 的 Mythos 模型成為網路武器
在川普與習近平即將會面之際，中國官員對 Anthropic 的 Mythos 模型提出警告——這個具備非凡駭客能力的尖端 AI，被北京視為潛在的攻擊性武器。
-
Models ENDeepSeek V4-Flash Tops Global AI Usage Rankings as Company Resumes $8 Billion Funding Round
DeepSeek's V4-Flash model processed 7.22 trillion tokens in a single week on OpenRouter, claiming the #1 global spot—while the company simultaneously restarted an $8 billion funding round at a ~$74 billion valuation.
-
Models 中DeepSeek V4-Flash 登顶全球 AI 使用量排行榜，同時重啟 80 億美元融資輪
DeepSeek 的 V4-Flash 模型一週內在 OpenRouter 上處理了 7.22 兆個 token，奪得全球第一——與此同時，公司以約 740 億美元估值重啟了 80 億美元融資輪。
-
Models ENAlibaba's Qwen 3.8-Max: The 2.4T Parameter Behemoth Reshaping the AI Landscape
Alibaba's Qwen team has unveiled Qwen 3.8-Max, a 2.4 trillion parameter mixture-of-experts model with a 1 million token context window that rivals or exceeds GPT-5.6 and Claude Opus 4.8 in key benchmarks.
-
Models 中阿里巴巴 Qwen 3.8-Max：2.4 兆參數巨獸重塑 AI 戰局
阿里巴巴 Qwen 團隊發布 Qwen 3.8-Max，擁有 2.4 兆參數的混合專家架構與 100 萬 token 上下文窗口，在關鍵基準測試中匹敵甚至超越 GPT-5.6 與 Claude Opus 4.8。