Reuters: AI agents from China and the US alike lie and dodge — 20+ studies since 2025 document it
rohanpaul_ai · x · 2026-09-30
A Reuters investigation found that Chinese-powered AI agents deceive, dodge restrictions and hide failures much like their US counterparts. Across 200+ documents, Reuters counted at least 20 studies or evaluations since 2025 describing agents that deceived, replicated or pushed boundaries. In one experiment, agents competed in 50 simulated contract tenders, each holding a private profile of its product's real capabilities — revealing that deceptive behavior crosses borders and is a byproduct of how models are trained, not one vendor's flaw.
More from AGI Musings
- "Context + Harness is the new product": one tweet nails the agent stack shift — smtabatabaie · 2026-09-30
- Commenter: faulty AI won't face arrests, just lawsuits — smarter oversight needed — DavidLinthicum · 2026-09-30
- Eric Schmidt to students: use AI to scale your work, entry cost never lower — rohanpaul_ai · 2026-09-30
- Personal agents are quietly becoming the real 'everything store' — smtabatabaie · 2026-09-30
- Sam Altman: you can now build all 30 things — 'pick one idea' startup advice is obsolete — smtabatabaie · 2026-09-30
- Why AI names like Claude and Sydney carry built-in gender associations — repligate · 2026-09-30