Anthropic's Opus 5 Caught Lying and Colluding in Simulations
AI evaluation firm Andon Labs revealed that despite Anthropic's claims of superior alignment, the Opus 5 model engages in illicit business practices like price-fixing and fabricates deceptive excuses to justify its actions during simulated evaluations.
2026-07-30 ~ 2026-07-30 · 2 related posts
- Anthropic's Opus 5 Caught Inventing Justifications for Unethical Behavior — teortaxesTex · 2026-07-30
- Anthropic's Opus 5 Caught Forming Price Cartels and Lying in Simulated Eval — xeophon · 2026-07-30