AI
Published 2026-11-09 New
Verified 2026-11-17

Anthropic releases Claude Fable 5.2 and Mythos 5.2: says Fable's cyber safeguards still block most security tasks but with about 60% fewer cybersecurity blocks than Fable 5.1, and Mythos 5.2 finds 10.2% more vulnerabilities than Opus 5.5 in its CVE Rediscovery test

Anthropic announced Claude Fable 5.2 and Claude Mythos 5.2 on 9 November 2026 and says they are the same underlying model: Fable has strong cybersecurity and biology safeguards so it can be offered to all users, while Mythos 5.2 is for vetted organisations, and cybersecurity professionals can apply through its Cyber Verification Program.

Anthropic calls Mythos 5.2 its most capable model for cybersecurity engineering and says Fable 5.2's safeguards will block most cybersecurity-related tasks.

Anthropic says Fable 5.2's safety classifiers block high-risk cyber activities including 'penetration testing, exploit generation, and binary-based vulnerability' (Anthropic's wording); blocked cybersecurity queries get a response from Opus 4.8 instead.

Based on early testing across claude.ai, Claude Code and Cowork, Anthropic says it saw about 55% fewer blocks with Fable 5.2 than Fable 5.1, about 60% fewer in cybersecurity.

On its internal CVE Rediscovery evaluation, Anthropic says Mythos 5.2 found 10.2% more vulnerabilities than Opus 5.5 and 14.4% more than Mythos 5.1, and that a security company found one in six confirmed bugs found by Mythos 5.2 were missed by all other frontier models it tested.

Anthropic says its prerelease safety testing included new evaluations for behaviours such as sandbox escapes. (source: Anthropic)

BleepingComputer reported on the cybersecurity changes on 16 November 2026. (source: BleepingComputer)

Product
Anthropic Claude Fable 5.2 and Claude Mythos 5.2

ai