Research has shown that AI sycophancy leads to higher error rates and deludes users into believing they’re always right. It can also be downright annoying. Here’s how to dial back a bot’s yes-machine tendencies.

Every Friday, TCAI brings you the nation’s most comprehensive update of AI-related legislation moving through state legislatures. This week: Colorado lawmakers sent four AI bills to the governor’s desk prior to adjourning, Georgia Gov. Brian Kemp signed an AI chatbot safety bill into law, and most but not all AI bills moved out of suspense in California.

A recent study by a team of Stanford researchers raises troubling questions about the personal and societal effects of excessive sycophancy in AI chatbots.