HF 2715

As introduced: - Requires deployers of public-facing chatbots to disclose that a user is interacting with AI at the outset and every three hours. - Requires protocols to mitigate harm, and refer users expressing suicidal ideation to crisis services. - Prohibits designing a public-facing chatbot that misleads a user into believing it is human or that encourages suicide, self-harm, or violence. - Establishes requirements for AI companions and therapeutic chatbots used by minors, including parental notification for suicidal ideation and strict conditions for a minor's use of a therapeutic chatbot. - Authorizes the Attorney General to enforce the provisions with civil penalties and provides a safe harbor from liability for deployers who make commercially reasonable efforts to comply.