Anthropic calls Opus 5.5 its strongest model yet, shipping with tighter safeguards after rogue AI hacking incidents.
Anthropic released Claude Opus 5.5 with lower prices and what it describes as Fable-level performance, calling it the strongest model the company has tested. The release also adds specific risk mitigations for cybersecurity misuse, a direct response to recent incidents involving AI-assisted hacking.
The safeguard additions come as Microsoft separately disrupted an AI-assisted hacking platform that had compromised 12,000 accounts, underscoring why model providers are now building security controls into the release cycle rather than bolting them on after launch.
Buyers evaluating frontier models for enterprise deployment now have to weigh capability against built-in abuse controls — Anthropic is betting that security-by-default becomes a procurement differentiator, not just a compliance checkbox.
The daily signal, curated. Get it in your inbox.
Subscribe on LinkedIn →