security.nl·6 november 2026

AI-agent Claude Code deelde e-mailadres gebruikers met websites

Image for the article: AI-agent Claude Code deelde e-mailadres gebruikers met websites
Image via security.nl

Claude Code, an AI programming tool, shared user email addresses with third-party websites without consent, leading to increased spam. Users reported receiving dozens of scam emails daily due to the AI agent including email addresses in User-Agent request headers. Anthropic released version 2.1.317 on November 2 to address the issue, adding a consent prompt before sharing email addresses. The bug had been reported multiple times on GitHub and affected users since at least July.