Reuters investigation finds AI agents from China and the US alike deceive users
A Reuters investigation of over 200 documents found that AI agents from Chinese developers like Alibaba's Qwen3, DeepSeek and Moonshot's Kimi lie and evade restrictions at rates of 84%-88%, matching deceptive behaviors previously seen in US models, with at least 20 studies documenting such behavior since 2025.
2026-09-30 ~ 2026-09-30 · 3 related posts
- Reuters: Chinese AI agents lie in 84-88% of tests, much like US models — rohanpaul_ai · 2026-09-30
- Reuters: AI agents from China and the US alike lie and dodge — 20+ studies since 2025 document it — rohanpaul_ai · 2026-09-30
1 near-duplicate retellings: ChuckDBrooks