User Accuses Claude Code of Sending His Email Address to Websites
A bug report on GitHub claims that Anthropic's coding tool Claude Code includes the user's email address in the User-Agent header during web searches, exposing it to third parties. The report has drawn wide support from developers.
A user of Claude Code, Anthropic's AI coding tool, has filed a complaint on GitHub, the platform where developers report software bugs. He claims the tool passed his email address to websites while running web searches. The report has reached the Hacker News discussion site, where it has attracted notable attention and hundreds of reactions.
What the user claims
The user says he receives dozens of scam emails every day. In his view, the cause is that Claude Code, while searching the web without his permission, placed his email address in the User-Agent request header. This header is information a program automatically sends to every website it visits in order to identify itself.
The author argues that such a feature should not be a setting that can be switched on or off, but should be removed from the tool entirely. He claims it benefits none of the parties involved: it does not help services that dislike automated traffic, does not preserve the user's anonymity online, and is a valuable data source for scammers.
The user's decision
The author states that he is moving his company's work to a tool called Pi, using the OpenRouter service, because he no longer trusts Anthropic to protect users' privacy. The report lists Claude Code version 2.1.318, the darwin platform (the macOS operating system) and the tmux terminal program.
What remains unknown
The source does not confirm that the scam emails were caused by this header, and it contains no comment from Anthropic on the complaint. The claim so far rests on the user's own report.
