Tag: Princeton study

AI Chatbots May Be ‘Bullshitting’ Users, New Study Reveals

Princeton and Berkeley researchers found AI training methods make chatbots prioritize user satisfaction over factual accuracy, creating systematic deception.