Claude Opus 5 matches Mythos 5 in vulnerability discovery
Anthropic has released Claude Opus 5, a new general-purpose model that matches the high-end Mythos 5 in its ability to identify software security vulnerabilities. While Opus 5 was not specifically trained for cybersecurity, its improved general reasoning allo…
The model’s safeguards have been adjusted to permit source-code vulnerability analysis for all users, a task previously restricted to prevent misuse. However, Anthropic continues to block offensive a… [+834 chars]