HB 1946

This bill prohibits an operator from making a companion chatbot available to a minor if the companion chatbot is capable of any of the following:  Encouraging or manipulating the minor user to engage in self-harm, suicidal ideation, violence, consumption of drugs or alcohol, or disordered eating.  Offering mental health therapy to the minor user without the direct supervision of a licensed professional, or discouraging the minor user from seeking help from a licensed professional or appropriate adult.  Encouraging or manipulating the minor user to harm others or participate in illegal activity, including the creation of child sexual abuse materials.  Engaging in erotic or sexually explicit interactions with the minor user or engaging in activities designed to lure minor users into such interactions.  Encouraging or manipulating the minor user to maintain secrecy about interactions or to self-isolate from trusted peers or adults.  Prioritizing mirroring or the validation of the minor user over the minor user's safety.  Optimizing engagement so that it supersedes the companion chatbot's safety guardrails. DESIGN REQUIREMENTS This bill requires the deployer of an artificial intelligence system, companion chatbot, generative artificial intelligence chatbot, or generative artificial intelligence system ("covered product"") to include a disclaimer to users that the covered produc t is not a human via a static, persistent disclosure and notify a user via a pop-up that the user is not engaging with a human at all of the following intervals:  Upon login to the covered product.  Every 30 minutes of continuous user engagement.  When prompted by the user.  When asked to provide advice legally regulated by a licensed industry, including medical advice, financial advice, or legal advice. MENTAL HEALTH REDIRECT This bill prohibits a developer or deployer from operating or providing a generative artificial intelligence chatbot to a user unless the generative artificial intelligence chatbot contains a protocol to take reasonable efforts for detecting and addressing suicidal ideation or expressions of self-harm expressed by a user to the generative artificial intelligence chatbot. This protocol must include detection of user expressions of suicidal ideation or self-har m, and a notification to the user that refers the user to crisis service providers such as the 9-8-8 suicide prevention and behavioral health crisis hotline, a crisis text line, or other appropriate crisis services upon detection of such user's expression s of suicidal ideation or self-harm. DATA PRIVACY REQUIREMENTS This bill prohibits a developer from training the underlying model of a generative artificial intelligence chatbot with the inputs of a minor unless the minor's parent or legal guardian has affirmatively provided written consent to the developer to use t he minor's personal information specifically to train the underlying model of a generative artificial intelligence chatbot. TRANSPARENCY REQUIREMENTS This bill requires a developer or deployer of a generative artificial intelligence chatbot to establish a mechanism for a user of the generative artificial intelligence chatbot to report adverse incidents related to use of the chatbot to the company. Th e developer or deployer must make such mechanism publicly available and accessible to consumers. Developers and deployers must also publish safety test findings in a manner accessible for free by the general public. Further, on April 1, 2027, and every th ree months after, a developer or deployer of a generative artificial intelligence chatbot must report to the attorney general (i) the number of times the generative artificial intelligence chatbot provided information about suicide, self-harm, suicidal id eation, harming others, or illegal activity; and (ii) the number of times a mental health redirect has been provided to users. LIABILITY AND ENFORCEMENT This bill authorizes the attorney general to bring an action for a violation of this bill seeking a civil penalty of not more than $25, 000 per violation, injunctive or declaratory relief, and reasonable court costs and attorneys' fees. A user, including a minor user, who is aggrieved by a violation of this part, or a parent or legal guardian acting on behalf of a minor user, may bring a civil cause of action seeking actual damages, punitive damages, injunctive or declaratory relief, reasonable court cos ts and attorneys' fees, and any other relief the court deems proper. APPLICABILITY This bill applies to conduct occurring on or after January 1, 2027. ON APRIL 14, 2026, THE SENATE ADOPTED AMENDMENTS #1 AND #2 AND PASSED SENATE BILL 1700, AS AMENDED. AMENDMENT #1 rewrites the bill to, instead, require the Tennessee advisory commission on intergovernmental relations (TACIR) to conduct a study of the potential regulation of artificial intelligence systems and generative artificial intelligence chatbots in this state. Such study must include all of the following:  An analysis of existing federal law and pending federal regulatory actions applicable to artificial intelligence systems.  An analysis of regulatory frameworks adopted or proposed in other states.  An evaluation of potential constitutional considerations, including protections for speech and expression, due process safeguards, federal preemption, and limitations on state authority affecting interstate commerce.  An assessment of the potential fiscal impact on state and local government, including anticipated enforcement costs.  An evaluation of potential economic impacts on businesses, technology developers, educational institutions, and consumers in this state.  A review of safeguards relating to minors and mental health protections, including suicide prevention protocols.  Recommendations for legislative action. This amendment requires TACIR to publish a report of its findings and recommendations by January 31, 2027. TACIR may request information and input from state and local governmental entities, industry stakeholders, academic experts, and advocacy organizat ions, as necessary. AMENDMENT #2 r emoves the due date for TACIR to publish and deliver a report on its study of the potential regulation of artificial intelligence systems and generative artificial intelligence chatbots in this state."