Anthropic launches Claude Security powered by Claude Opus 4.7
Anthropic just dropped Claude Security, a cybersecurity tool powered by the Claude Opus 4.7 model. Right now it's in public beta and available for Enterprise users.
The tool scans code for vulnerabilities and even suggests patches, all as part of Anthropic's bigger push to help spot risks in open-source software (think: Project Glasswing).
Claude security uses multi stage validation
Claude Security uses a multi-stage validation system that double-checks its findings and gives each issue a confidence rating, so developers know what to fix first.
It also generates clear reports on how serious a bug is and how to patch it up.
You can schedule regular scans or dismiss issues with notes for easy tracking.
Plus, Anthropic is teaming up with CrowdStrike to make sure its AI fits right into cybersecurity platforms, and Deloitte is deploying Claude Security to help enterprises strengthen their security posture.