Sycophancy

AIAI that agrees too much may be hurting human judgment

Study finds AI chatbots over-affirm users, reinforcing harmful beliefs and reducing accountability in personal decisions.

Joseph Shavit