Anthropic announces Claude Security to find and fix code vulnerabilities
Anthropic announced Claude Security, a new AI-powered tool designed to help companies find and fix security issues in their code.
Currently in public beta for enterprise-tier users, it uses the Claude Opus 4.7 model to scan codebases and suggest targeted patches.
This follows Anthropic's earlier Project Glasswing, which works with big names like AWS, Microsoft, and Google to protect open-source software.
Claude security scans with confidence ratings
Claude Security can check entire repositories or specific folders for vulnerabilities, giving each finding a confidence rating so teams know what to tackle first.
It also offers scheduled scans, easy ways to export results, and tools for dismissing or documenting issues.
To make things even stronger, Anthropic is teaming up with cybersecurity leaders like CrowdStrike and Palo Alto Networks to integrate its tech into their platforms.