Anthropic just caused chaos across the developer community after a bungled DMCA takedown campaign accidentally nuked thousands of unrelated GitHub repositories. The AI safety company, maker of the Claude assistant, was scrambling to remove leaked source code when its overly aggressive copyright notices took down repos that had nothing to do with the leak. Within hours, developers across the platform found their projects suddenly inaccessible, sparking immediate backlash before Anthropic executives admitted the mistake and started retracting the notices.

The developer community woke up to an unexpected crisis Wednesday when Anthropic unleashed a wave of DMCA copyright takedown notices that swept across GitHub like a digital wildfire. What started as a legitimate effort to contain leaked proprietary code turned into a mass deletion event that caught thousands of innocent developers in its crosshairs.

According to reports flooding social media and developer forums, repositories spanning everything from personal projects to open-source tools suddenly disappeared without warning. The common thread? Anthropic’s legal team had flagged them as containing stolen intellectual property, but the vast majority had zero connection to the AI company’s codebase.

The leak itself appears to have originated from an internal security breach at Anthropic, though the company hasn’t disclosed specifics about how its source code ended up in the wild. What’s clear is that once executives discovered the leak, they moved fast – perhaps too fast – to scrub it from the internet’s most popular code repository platform.

GitHub, owned by Microsoft, processes DMCA takedown requests through a semi-automated system designed to protect copyright holders while maintaining platform integrity. But when Anthropic submitted what sources describe as an overly broad list of allegedly infringing repositories, GitHub’s system apparently lacked sufficient guardrails to catch the errors before executing the removals.

The blowback was immediate and fierce. Developers took to Twitter and Hacker News to share screenshots of takedown notices, many expressing confusion about how their projects could possibly contain Anthropic’s proprietary code. Some pointed out they’d never even used Claude or any Anthropic API in their work. Others noted their repositories predated Anthropic’s founding in 2021.