Anthropic released Claude Security in public beta for Enterprise customers. The tool evolved from a research preview previously known as Claude Code Security. It enables cybersecurity teams to scan entire codebases for vulnerabilities and generate potential fixes.
Anthropic’s Opus 4.7 model powers the new security tool. The model analyzes code by tracing data flows and component interactions to mimic human researchers. This capability identifies complex flaws that traditional pattern-matching tools often miss.
The launch intensifies competition within the AI-powered cybersecurity market. Partners including CrowdStrike have already integrated the new model into their platforms.