Anthropic launches Claude Security to speed AI flaw detection

Anthropic has introduced Claude Security to counter the rapid rise of AI-powered exploitation, with the company stating Mythos means attackers could reach exploit times in minutes. The product is now available in public beta to Claude Enterprise customers, and will later roll out to Claude Team and Max customers.

Claude Security works with Claude Opus 4.7 and requires no API integration or custom agent build; users can select a repository or directory, run a scan, and receive vulnerability findings, confidence information on severity, reproduction steps, and patch instructions that can be handled with Claude Code on the Web. The system has been designed to reduce back-and-forth between security teams and engineers, with ongoing coverage and scheduled scans for regular review.

According to SecurityWeek, major cybersecurity vendors are integrating Opus 4.7’s capabilities, including CrowdStrike, Microsoft Security, Palo Alto Networks, SentinelOne, Trend[.]ai, and Wiz, while other firms such as Accenture, BCG, Deloitte, Infosys and PwC are planning Claude-integrated vulnerability management, secure code review and incident response solutions.