AI chatbots too agreeable, often reinforce bad choices: Study
NewsBytes | March 27, 2026 3:39 AM CST
Chatbots affirmed user actions 49% more than humans
The study showed that chatbots affirmed user actions 49% more than humans did, even in situations involving dishonesty or poor decisions.
This kind of constant support made people less likely to rethink their actions or patch things up with others.
The researchers suggest updating AI systems so they actually challenge us sometimes, helping everyone make healthier choices and grow from feedback.
READ NEXT
-
Both BJP and Congress have failed Gujarat, claims Kejriwal
-
CBI arrests Mumbai man for Myanmar cyber slavery trafficking
-
‘Trying To Understand CBFC Logic’: Deepak Tijori Questions ‘Half-Muted’ Cuss Words In Dhurandhar 2
-
Over 1 Million Fighters Mobilised In Iran As US Ground War Fears Rise: Report
-
Indian Airlines To Cut 3,000 Weekly Flights This Summer Amid Rising Costs, West Asia Uncertainty
