AI bill targeting chatbot risks, child safety passes CT state Senate

The Connecticut state Senate passed a bill on Wednesday that aims to regulate AI chatbots, with a focus on risks to children and expanded protections for whistleblowers. The legislation requires AI providers to review and mitigate potentially harmful interactions between chatbots and minors, including content that promotes self-harm, bullying, or grooming. It also establishes a legal framework for employees in the AI industry to report 'catastrophic' risks without retaliation, creating a new whistleblower category under state law. Supporters of the bill, which passed with bipartisan support in the Senate, argue it addresses gaps in existing consumer protections as AI tools become more integrated into everyday life. The bill now moves to the state House of Representatives for further consideration. Connecticut is one of several states considering AI regulation in the absence of comprehensive federal legislation, with previous efforts including a separate bill targeting AI-generated deepfakes in political elections. The 'catastrophic risk' threshold for whistleblower protections is defined in the legislation as a harm that would result in mass casualties or severe, widespread damage to public infrastructure or the economy.