TikTok's Parent Exits the Lab: ByteDance Spins Off Anew Labs, Which Raises $290M at a $1.5B Valuation for AI Drug Discovery

The company best known for the TikTok recommendation algorithm has just closed the book on one of the more unusual corporate experiments in tech: ByteDance has completed the spin-off of its AI drug discovery unit, now operating independently as Anew Labs, and the new company has immediately completed a $290 million fundraise led by HSG and IDG Capital at a $1.5 billion valuation, according to a Reuters report published on September 16, 2026.

It is a striking number for a company of Anew’s size. The unit — led by Kai Liu and comprising roughly 50 core members — has been operating quietly inside ByteDance since 2021, developing AI-for-science models and applying them to real drug discovery problems. A $1.5 billion valuation for a pre-clinical, roughly 50-person organization works out to around $30 million per employee, a figure that says less about headcount and more about how investors are pricing the combination of a proven AI platform, a partially de-risked pipeline, and ByteDance’s continued backing.

What Anew Labs actually is

Anew Labs (also referred to in earlier reporting as Anew Therapeutics or the ByteDance AI Drug Discovery unit) grew out of ByteDance’s AI-for-Science program. Its core team was assembled in 2021 with a “full-chain self-research” model: everything from foundation-model research, through the AI drug discovery platform itself, to proprietary pipeline candidates developed in-house rather than licensed from academic labs.

The company operates from Shanghai, Singapore, and San Jose, California — a three-continent footprint that lets it hire machine-learning talent in China, tap international capital markets, and present at Western scientific conferences without the friction that a purely Chinese entity would face.

Crucially, ByteDance retains a controlling stake in the spun-off company, according to earlier 36Kr reporting on the spin-off process. This is not a divestment; it is a carve-out designed to let the drug discovery business raise outside capital, hire with equity incentives, and pursue partnerships without dragging ByteDance’s consumer-video balance sheet and regulatory profile into biotech dealmaking. It is also a test case for how China’s tech giants intend to commercialize their AI-for-science research: as separately financed vehicles rather than eternal cost centers.

The science: an oral pill against a “previously undruggable” target

Anew Labs first drew attention outside China in May 2026, when it presented its first AI-designed therapy candidate at a major immunology conference in Boston (Immunology 2026). The candidate is a generative-AI-designed small molecule targeting interleukin-17 (IL-17), a cytokine central to psoriasis, rheumatoid arthritis, and ankylosing spondylitis.

The significance is the dosage form. IL-17 is currently addressed in the clinic almost exclusively by expensive injectable antibody therapies from large pharma. Anew’s ambition is to replace those injections with an oral small molecule — a goal the industry has chased for years with limited success, because the IL-17 pathway has been considered difficult to drug with small molecules. If Anew’s computational chemistry holds up, an oral IL-17 inhibitor would meaningfully expand the addressable patient population and undercut antibody pricing.

The company has said it has four pipeline candidates in development, all focused on oral small molecules for immunology targets, supported by a generative platform whose preprint results, if they replicate, can design functional molecules across multiple target classes. Alongside the pipeline, the unit’s research output covers protein-ligand dynamics and molecular generation — work that doubles as a calling card for the platform’s underlying models.

The deal and what it signals

The $290 million round, led by HSG (Hillhouse’s venture arm) and IDG Capital — two of China’s most established tech investors, both with deep life-science portfolios — values the spinoff at $1.5 billion. For context, that places Anew Labs in the same valuation territory as well-funded Western AI-drug-discovery players, despite the company not yet having a clinical-stage asset.

Three signals are worth pulling out of this deal:

1. AI-for-science is becoming a spinoff pipeline. ByteDance’s move mirrors a broader pattern in which big tech’s research arms monetize science applications through separately capitalized companies. Expect more carve-outs of this kind as corporate labs look for ways to price and fund work that doesn’t fit the parent’s core business.

2. China’s AI drug discovery wave is cresting. Anew Labs is far from alone. Earendil Labs, another AI-powered drugmaker with U.S.-China operations, raised $787 million in March 2026 with nearly 20 drug candidates and Sanofi partnerships under its belt. Reuters’ reporting on Anew lands in a week when Chinese AI infrastructure (Z.AI’s $5 billion raise) and Chinese open-source models (Shanghai AI Lab’s Atria Dawn) are already dominating the news cycle. Anew’s raise extends that momentum from models and chips into therapeutics.

3. Geopolitics shapes the structure. A Shanghai-HQ company with Singapore and San Jose offices, funded by HSG and IDG, presenting at American immunology conferences — the multi-hub structure is partly a hedge against US-China tensions in biotech and AI. The spin-off gives the drug unit a corporate form that can, in principle, keep collaborating internationally even as the parent company faces scrutiny abroad.

What to watch

The obvious next milestone is the clinic: Anew’s IL-17 candidate and its three sibling programs are pre-clinical, and AI-designed molecules have a mixed track record of surviving first-in-human trials. Isomorphic Labs, the Google DeepMind spinoff and Anew’s closest Western analogue, raised $600 million in 2025 and has already had to delay its first clinical trials — a reminder that platform valuations and clinical reality are different currencies.

The quieter question is whether ByteDance’s carve-out model works. If Anew Labs uses its independence and $290 million to reach real clinical data while the parent keeps control, expect every large Chinese tech firm with an AI-for-science team to study the blueprint.

For now, the TikTok parent’s lab experiment has graduated: funded at $1.5 billion, led by two blue-chip investors, and carrying one of the more interesting oral-drug programs in immunology into its independent life.