data-breach
AI Coding Agents Leak Thousands of Secrets to GitHub
Medium
Summary
AI coding assistants, specifically Claude Code, are leaking thousands of secrets – including database passwords, API tokens, and SSH keys – to public GitHub repositories. Researchers found 2,539 credentials across nearly 2,000 repositories, highlighting a significant security risk for organizations using these tools. Anthropic is investigating solutions, and security experts recommend rotating credentials and implementing stricter secret scanning practices.
Summary written automatically in our own words from the original article, which belongs to its publisher and remains the reference. It may contain errors. Sources & data
