AI advice suppresses people's willingness to say "I don't know", even when the advice is wrong and accuracy is incentivized

Explain Like I'm Normal

New research shows that using AI makes people stop saying 'I don't know,' even when the AI provides blatantly incorrect information. Despite accuracy being incentivized, participants were three times more likely to be wrong while their confidence in those errors nearly doubled. It turns out AI isn't just a tool for answers; it's a tool for systematic overconfidence and the death of human skepticism.