X Shuts Down Nitter With Cease-and-Desist — and the Open-Source Code May Be Next

After seven years of letting anyone read public X posts without an account, Nitter is gone — and the legal letter that ended it is now aimed at the project’s source code itself.

On August 24, 2026, X Corp. sent cease-and-desist letters demanding the permanent shutdown of Nitter’s instances and the removal of the project’s code repository. The flagship Nitter.net instance went dark ahead of X’s 5 p.m. EST deadline on August 25, and the project’s maintainer, known as Zedeus, paused development while seeking legal counsel. The related service XCancel, which relied on Nitter’s codebase, has also posted about receiving its own cease-and-desist letter.

What the letters allege

X’s lawyers accuse Nitter of violating platform rules through scraping data, accessing X accounts and session tokens, and what the letter calls “unlawful use and circumvention” of X’s API. The legal grounding rests on two statutes: the Texas Harmful Access by Computer Act — specifically Section 33.02 of the Texas Penal Code, which criminalizes accessing computer systems without the owner’s “effective consent” — and the Lanham Act, the federal trademark statute more commonly associated with brand disputes.

The Register noted the irony that the letters came from a firm headed by a self-described free speech absolutist, a pointed reference to Elon Musk’s long-standing positioning of X as the platform for unfettered expression.

Why this matters beyond privacy fans

Nitter’s original pitch was simple: a lightweight, ad-free, JavaScript-free frontend for reading public X posts. It stripped out tracking cookies and ads, and required no login. That made it a favorite not just of privacy-conscious users, but of a less visible constituency — security researchers publishing findings, open-source projects like GrapheneOS posting announcements, and increasingly, AI agents and automated research tools that needed programmatic access to public post content.

That last group is the quiet casualty here. As The New Stack observed, the move tightens X’s stance against third-party clients and scraping infrastructure “that AI agents and researchers have relied on for public post access.” In an era where agentic AI systems routinely fetch web content to answer questions, verify claims, and monitor breaking events, the shrinking number of open windows into major social platforms concentrates access in the hands of those willing to pay for official APIs — or to strike licensing deals.

Nitter itself had a complicated relationship with X’s technical barriers. The service originally died in 2024 when X shut off the guest-access mechanism it depended on, taking Nitter.net offline at the time. It later returned by using real X accounts to fetch posts, according to the project’s GitHub repository. That workaround kept the service alive but placed it on precarious footing: X controlled those accounts and could change the rules around them at any moment. Those account-based access methods are now the centerpiece of X’s legal theory.

The open-source code question

The most legally interesting aspect of X’s demand is that it targets the repository itself, not just the running service. Because Nitter is open source, anyone can fork the code and run an independent instance — taking Nitter.net offline doesn’t eliminate Nitter. X’s demand that the code be removed entirely is an attempt to change that.

But removing code from GitHub is a higher bar than sending a letter. Under Section 1201 of the Digital Millennium Copyright Act, software can be targeted if it’s used to bypass technology protecting access to copyrighted material — even if the software doesn’t contain that material itself. Yet a mere claim that software circumvents a technical restriction isn’t sufficient. Per GitHub’s own guidelines for circumvention claims, a complainant must identify the copyrighted material being protected and explain precisely how the code circumvents the technology controlling access to it.

The precedent here cuts both ways. In 2020, GitHub removed the YouTube-dl project following a Section 1201 complaint from the Recording Industry Association of America — then later restored the repository and reformed its process, adding technical and legal review before pulling code when a circumvention claim isn’t clear-cut. For now, the Nitter repository remains on GitHub in an archived, read-only state. The code can still be viewed and forked, even though its developer is no longer working on it.

Platform control vs. the open web

Strip away the legal particulars and the story is about the asymmetry of platform power. Open source gives developers control over code, but not over the platform that code runs against. X still owns the service, the accounts, the session tokens, and the API — and it has now demonstrated that its control extends beyond rate limits and paywalls to demanding the erasure of the tooling itself.

It’s also a data point in a broader trend. Social platforms have spent the past two years walling off their content from scrapers and crawlers, with AI companies’ voracious training-data appetite as the stated justification. X’s paid API tiers, Reddit’s licensing deals, and publisher lawsuits against AI firms all point in the same direction: public content is becoming progressively less public, and the enforcement mechanism of choice is increasingly legal rather than technical.

For the AI ecosystem, the practical effect is a shrinking commons. Agents that could once read a public post through a neutral frontend now face a starker choice: authenticate through an official, metered API; rely on licensed intermediaries; or go without. Researchers studying misinformation, disaster response, and public discourse — work that historically leaned on open access to platform data — lose another instrument.

Nitter may well return in some forked form; the code is out there, and the archival instinct of the open-source community is strong. But X’s willingness to reach past a running service and aim at the source code itself sets an uncomfortable precedent for every tool that interoperates, however modestly, with a closed platform.

Seven years of ad-free, account-free access to public posts ended not with a technical change, but with a letter. What replaces it — for researchers, for privacy-minded readers, and for the AI agents that quietly read the web on our behalf — is a question the industry will be wrestling with long after Nitter’s farewell page loads its last time.