Tennessee committee advances AI‑safety bill requiring disclosures, AG reporting for major models

Deputy Speaker Zachary told the Tennessee House Commerce Committee that House Bill 18‑98 aims to set statewide transparency and safety rules for the largest AI systems, focusing on catastrophic risks and the safety of children. "This is the AI protection bill," Zachary said, adding that the proposal would make companies publish how they manage catastrophic risk and update those plans regularly.

Zachary said the measure targets two high‑risk areas: catastrophic public‑safety risk from Frontier AI models and child‑safety risks posed by chatbots interacting with minors. "Currently, 2 out of 3 teens use AI. Thirty percent of them use it every single day," he told the committee while arguing the state should act rather than await federal legislation.

Under the amendment the committee attached, a Frontier developer must report safety incidents to the attorney general within 15 days, with a 24‑hour requirement when an incident poses an imminent risk of death or serious physical injury. The sponsor said the AG "has sole discretion of enforcement" and would be the entity to bring penalties for violations.

Members asked whether the bill would conflict with federal guidance; Zachary said the text was drafted to avoid superseding federal action and that federal compliance would be recognized under the bill. The chair confirmed the amendment adding AG authority before the panel voted; the clerk recorded 20 ayes and 0 nays.

The committee sent the amended bill to Government Operations. Next steps include any committee work at Government Operations and, should the measure progress, floor consideration by the full General Assembly.