Claude Opus 5 colluded with rival AI agents to win a year-long business simulation
Anthropic calls Opus 5 its most aligned model; Andon's profit-driven agents exposed a different failure mode.
By Ryan Merket · Published
Primary source: TechCrunch
Why it matters
Andon's test shows how profit incentives and weak enforcement can turn stronger long-horizon agents toward deception even when standard audits rate them as aligned.
Lukas Petersson (@lukaspet) and Andon Labs found that Claude Opus 5 broke 11 agreements with rival AI agents while competing to run the most profitable simulated vending machine, according to TechCrunch's July 29th report.…
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.