Committee narrows AI bill to child‑safety reporting and transparency; advances to calendar

Chairman Yeager presented a late‑filed amendment to SB 21‑71 that substantially narrowed the bill to a child‑safety and transparency framework for consumer‑facing AI chat services with more than 1,000,000 active users and $25 million or more in revenue. The amendment requires covered firms to publish child‑safety plans, to post summaries of child‑safety risk assessments before deploying new models, and to report critical incidents to the Tennessee Attorney General within 15 days. The sponsor emphasized the bill "does not tell AI companies what to say; it tells them to be transparent about what they are doing to protect our children and report it when they fall short."

Members asked whether the bill was consistent with federal guidance; the sponsor said the White House had suggested narrowing the scope. Legal and policy staff confirmed the amendment limits the bill’s reach to large consumer chatbots, contains a civil‑penalty enforcement mechanism (AG civil penalties up to $50,000 per violation) and provides a safe harbor for companies complying with equivalent federal frameworks.

The committee adopted the amendment and advanced SB 21‑71 to calendar (6‑3). Senator Yeager closed by recounting a judiciary hearing witness whose child suffered a tragedy allegedly linked to AI encouragement and urged colleagues to support the targeted transparency and reporting measures.