Anthropic intentionally avoided training Opus 5 on cyber tasks, so while it has improved at finding vulnerabilities through general capability gains and comes close to Mythos 5 at detection, it remains substantially behind on exploitation — turning vulnerabilities into material threats. ✦ AI generated
Anthropic (release post) · Simon Willison's Weblog · 2026-07-24 · original ↗
As with its predecessor, Opus 4.8, we've intentionally avoided training Opus 5 on cyber tasks. The model has nevertheless improved substantially on these tasks as a result of becoming more generally capable, and it comes close to Mythos 5 at finding cybersecurity vulnerabilities. However, it remains substantially behind Mythos 5 on the exploitation of those vulnerabilities—that is, in turning vulnerabilities into material cyber threats.
Read full article ↗excerpt · fair-use quotation