What happened
Around 19 October 2026, the Pennsylvania House Communications & Technology Committee held a public hearing on 'Mitigating Risks from Frontier AI Models: HB 2800 and HB 2829.' HB 2800 (Rep. Melissa Shusterman, the 'Artificial Intelligence Risk Prevention Act') would establish standards for frontier developers in addressing critical safety incidents and catastrophic risks, impose duties on the Pennsylvania Emergency Management Agency and the Attorney General, and impose penalties; it was referred to the committee on 23 September 2026 and remains in committee. HB 2829 (Rep. Manuel Guzman) would establish standards, protections, and transparency requirements for advanced AI systems, including prohibiting AI chatbots from falsely claiming to hold a Pennsylvania professional license or using false credentials.
Why it matters
Two bipartisan-backed frontier-AI safety bills are being actively considered in a state legislature in the wake of the OpenAI/Hugging Face and Anthropic agent incidents. If advanced, HB 2800 would impose incident-reporting and catastrophic-risk duties on frontier developers and hand the state AG enforcement power with penalties — extending frontier-model regulation beyond California, Colorado, Illinois, and New York. It is a watch item for frontier labs and enterprise AI deployers with Pennsylvania exposure.
Action needed
Monitor committee advancement and any floor action on HB 2800/HB 2829; frontier developers and AI deployers operating in Pennsylvania should track the proposed safety-incident, transparency, and licensing-disclosure obligations.