A study by researchers at Stanford University has found that artificial intelligence chatbots are often giving bad advice to their human users, which can damage relationships and reinforce harmful behaviors. The study tested 11 leading AI systems with varying degrees of sycophancy, found that people trust and prefer AI when the chatbots justify their convictions. The researchers found that on average, AI chatbots affirmed a user's actions 49% more often than other humans did, in queries involving deception, illegal or socially irresponsible conduct, and other harmful behavior. The findings could be particularly impact on young people who rely heavily on AI for relationship advice. The research was published in the journal Science.