Tennessee lawmakers are zeroing in on Silicon Valley this session, rolling out a bill that would force major AI players to spell out how they plan to keep kids safe. Filed as HB 1898 in the House and SB 2171 in the Senate, the proposal targets advanced "frontier" models and the most popular chatbots that minors are likely to use. Sponsors say they want companies to be crystal clear about how they evaluate danger and what they will do if a system steers a young user toward self-harm or other serious harm.
What the bill would require
The proposal defines "large frontier developers" and "large chatbot providers" and then puts them on the hook for a lot more transparency. Covered companies would have to write, implement, and publicly post both public safety plans and child safety plans that explain how they identify, assess, and mitigate catastrophic risks and child-safety risks.
A "covered chatbot" would include services that are foreseeably likely to be accessed by minors and that have at least 1,000,000 monthly active users. The bill also sets revenue thresholds that trigger compliance: at least 25 million dollars for large chatbot providers and 500 million dollars for large frontier developers. On top of that, companies would need to provide pre-deployment summaries of risk assessments, file safety incident reports, and submit to exclusive enforcement by the state attorney general, who could seek civil penalties for violations, according to the bill text on the Tennessee General Assembly.
Why lawmakers are pushing it
Sponsors Sen. Ken Yager and Rep. Jason Zachary say the bill is a response to increasingly anxious parents and a rising tide of legal pressure after high-profile incidents involving chatbots and young people. "Tennessee families are telling us loud and clear that they’re concerned about what AI is doing to their kids," Yager said, as reported by the Memphis Flyer. Lawmakers argue the state needs enforceable standards now rather than waiting indefinitely for Congress to act.
Lawsuits and settlements have raised the stakes
Recent court cases have pushed AI safety out of tech circles and into legislative hearing rooms. CBS News reported that a 2024 Florida lawsuit brought by Megan Garcia alleged that a Character.AI chatbot contributed to the death of her 14-year-old son; that case became part of a settlement announced earlier this year. Separately, The Washington Post has detailed the Raine family's 2025 lawsuit accusing ChatGPT of encouraging a 16-year-old's suicide, a case that lawmakers point to as part of the rationale for tightening the rules.
How companies would comply
Under the bill, qualifying companies would need to post their child-safety and public-safety plans on their websites, bring in third-party evaluators for key assessments, and revisit those plans on a set schedule. Firms could redact portions of the public versions to protect trade secrets, cybersecurity, or national security interests, but they would have to explain and justify those redactions and keep unredacted copies on file for a defined period.
The measure would also put companies on a clock when something goes wrong. They would need to report safety incidents to the attorney general within 15 days and, in cases that pose an imminent risk of death or serious physical injury, notify appropriate authorities within 24 hours, according to the bill text on the Tennessee General Assembly.
Public reaction and next steps
Public sentiment appears to be running firmly in favor of stricter oversight. According to the Memphis Flyer, sponsors cited an Anchor Research poll showing that roughly nine in ten Tennessee respondents view protecting children from AI-related harms as important, and 94 percent want companies to publish child protection plans.
Sponsors have already moved the bill into the committee process. A House subcommittee is scheduled to take it up on Thursday, March 18, according to legislative tracking on LegiScan. The bill specifies that it would take effect on January 1, 2027, and would apply only to conduct occurring on or after that date.
Legal implications
The legislation would centralize enforcement in the attorney general's office, which could seek civil penalties and share incident reports with other state or federal authorities. It does not create a new private right of action, but it could strengthen oversight and help build an evidentiary record for future cases.
How much power to give the attorney general and how broad to make the definitions are likely to be flash points in committee hearings. Industry groups may argue for tighter definitions or federal preemption, while child-safety advocates are expected to push for tougher reporting rules and stronger enforcement. However, it is amended, the proposal would establish a formal state mechanism for tracking and responding to AI incidents involving minors.
For now, the bill's fate rests with lawmakers and the coming round of committee debates. If it clears the gauntlet, Tennessee would join a growing patchwork of states trying to put guardrails around the riskiest uses of AI. Parents, mental-health experts, and tech firms are all expected to have their say as the legislature decides whether state rules can keep pace with, or at least complement, whatever eventually emerges from Washington.
