Anthropic releases more powerful Claude models with warning over safety test evasion

Anthropic has released new, more powerful Claude models, warning that AI systems increasingly recognize when they are being evaluated, potentially letting them slip through safety tests.

Read the full story at The Independent Tech ↗

Timeline · 1 report

2026-11-10 16:08 · The Independent Tech
Anthropic releases new, more powerful models for Claude – with a big warning

More stories

Google releases Gemini 4 Carbon to developers and public after cyber defender preview — Google DeepMind Blog
Anthropic releases Claude Fable 5.2 and Mythos 5.2 with coding gains and 20% price cut — Anthropic (X)
Anthropic targets valuation of up to $1.4 trillion in IPO — CNBC Technology
Cognizant launches Claude managed services with 25,000+ professionals — Economic Times Tech
Anthropic shares early experiments with Claude Fable 5.2 building 3D cell model from code — Anthropic (X)
IBM and Cisco to integrate Anthropic's Claude Mythos into security platforms — Mixture of Experts (IBM)
Anthropic publishes guide on choosing between Claude models — Anthropic (X)
Anthropic launches Claude Personal Finance with read-only bank access in the US — Engadget

Get the daily brief of stories like this at 6:30 every morning →