Though indefinitely postponed, the proposed bill would require public AI chat services to protect young users, respond to self-harm prompts, and disclose when people could mistake AI for a human.
Operators of covered public AI chat services would have to tell users under 18 that they are interacting with AI, using a notice that stays visible or repeated notices during a session.
For minors, operators would have to take reasonable steps to block explicit sexual images, certain sexual statements, and statements that make the AI seem human. They would also offer privacy and account tools to minors and, in some cases, their parents or guardians.
Operators would have to disclose when a reasonable person could mistake their AI for a human and set up responses that try to direct users who ask about self-harm to crisis help. They could not knowingly program claims that the service provides professional mental health care.
The Attorney General could sue an operator over a violation and seek damages or penalties, capped at $500,000 per operator. The bill would not let individuals sue on their own.
Operators would have to tell each minor account holder that they are interacting with AI. A minor is someone the operator actually knows, or has reasonable certainty, is under 18. The rule covers public services that mainly simulate human conversation, with listed exclusions for services such as developer tools, narrow-topic applications, and business-only systems. The bill would take effect July 1, 2027, if enacted.
Operators would have to take reasonable steps to prevent specified sexual or misleading outputs for minors and offer tools to manage privacy and account settings. These duties apply to minor account holders of covered public conversational AI services. The safeguards cover explicit sexual images, certain sexual statements, and statements that could make a reasonable person think the service is human, including simulated emotional dependence or adult-minor romantic role-play. Operators must offer tools to parents or guardians of users under 13 and, as appropriate based on relevant risks, to parents or guardians of older minors. The bill would take effect July 1, 2027, if enacted.
Operators would have to disclose when a reasonable person would be misled into thinking they are interacting with a human, set up a response protocol for self-harm prompts, and not knowingly program claims that the service provides professional mental or behavioral health care. The disclosure and self-harm protocol apply to users of covered public conversational AI services; the disclosure is required when a reasonable person would be misled into thinking they are interacting with a human. The self-harm protocol must include reasonable efforts to refer users to crisis services, such as a hotline or text line. The bill would take effect July 1, 2027, if enacted.
The Attorney General could sue an operator for a violation and seek relief, including damages, civil penalties, and legal expenses. The Attorney General could bring a case on behalf of Nebraska or a person harmed by a violation. Civil penalties would be at least $1,000 per violation and no more than $500,000 per operator. The bill would not create a private right to sue. It would not make a model developer liable for a third party's violation. The bill would take effect July 1, 2027, if enacted.
No hearings or votes scheduled.