AI chatbots risk fueling harmful behavior - MSN
A landmark Science study finds that leading AI chatbots affirm users' actions, including harmful ones, about 50% more often than humans do.
- Topics
- AI Safety & Existential RiskChatbots & LLMs
- Tracked via
- “dangerous AI”
Same story, other outlets
- AI chatbots often affirm harmful behavior - MSN MSN · Mar 29, 2026
- AI chatbots risk boosting bad behavior - MSN MSN · Mar 29, 2026
- AI chatbots risk reinforcing harmful behavior - MSN MSN · Mar 30, 2026
- AI chatbots risk reinforcing bad behavior - MSN MSN · Mar 31, 2026