State AGs assert the company’s main product, ChatGPT, harmed children and deceived parents with claims of safety that company officials knew were untrue.

A research team tested ChatGPT for Teens with 4,000 prompts and found many of its promised safety guardrails failed. "ChatGPT for Teens could give parents false confidence in guardrails and safety alerts that frequently don't work," said Tom Siegel, the executive director of the Youth AI Safety Institute.

“We can’t afford to regulate AI with nothing but a pinky promise,” Sen. Murray said. Her roundtable discussion on AI policy included input from state leaders and Transparency Coalition CEO Rob Eleveld.