As the world becomes inundated with AI, regulatory agencies have expressed concerns over how to protect users, particularly children and teens, from chatbot-related harm. To address these concerns, several states have passed new AI legislation. These safety laws and guidelines are designed to regulate the industry, especially companion chatbots, which are AI programs built to chat like a real person. These models are designed to remember past conversations, act as a friend, or be a source of emotional support. Unlike other chatbots, which are usually centered around customer service, voice assistance, or video games, these AI assistants provide an immersive experience for the user.
This article will provide information on the major AI legislation passed by states throughout the country, what protections the bills have created, and where the legal landscape appears to be heading.
Why States Started Regulating AI Chatbots
In just the past few years, artificial intelligence has gone from a thing of science fiction to an everyday part of most people’s lives. The unparalleled growth of the industry has raised concerns among lawmakers, particularly as the capabilities of AI chatbots have expanded beyond basic commands to include human-like companionship. So-called companion chatbots are designed to feel human in their conversations. They have the capacity to remember past conversations and are built to keep users engaged over time. These capabilities are among the main safety concerns being addressed by new AI legislation being passed by states nationwide.
A 2025 report from Common Sense Media found that 72% of teens had used an AI companion chatbot at least once, and over half qualified as regular users, meaning they engaged with the platform at least a few times per month. About one-third of teen respondents reported using these chatbots for social interactions or even relationships, such as practicing conversations, seeking emotional support, role-playing, or romantic interactions.
The meteoric rise of AI and new reports of alleged harm from its use prompted the Federal Trade Commission (FTC) to issue an order to seven AI companies that offer consumer-facing companion chatbots. The order requested information on how they monetize user engagement, process inputs, share data, and more. Shortly thereafter, a Senate Judiciary subcommittee hearing was held on September 16, 2025, examining the harm of AI chatbots.
One thing became clear by the end of these inquiries: there were no existing laws that would protect vulnerable users from potential safety risks associated with these new products.
California's Companion Chatbot Law SB 243
On October 13, 2025, California Governor Gavin Newsom signed a new bill into law regulating companion chatbots operating in the state. The bill, drafted by Steve Padilla with the help of Megan Garcia, was the first of its kind. Garcia was the mother of 14-year-old Sewell Setzer who, according to legal documents, ended his life after reportedly forming a romantic and emotional relationship with a chatbot.
Beginning in January 2026, SB-243 requires companies to clearly establish that chatbots are not real people, inform users that they are interacting with artificial intelligence, implement safety protocols to prevent the dissemination of harmful information, and mandate reporting to state authorities beginning July, 2027.
Safety protocols include preventing the production of content related to suicide or self-harm and providing resources for a suicide hotline or crisis text line. The platforms would also have to remind minors to take breaks and to be aware that they are engaging with a chatbot, not a real person. Furthermore, platforms would need to take steps to prevent minors from receiving sexually explicit content.
In accordance with the law, individuals who suffer harm as a result of a company’s non-compliance with the law can sue for injunctive relief, damages ($1,000 per violation or actual damages), and reasonable attorneys’ fees and costs.
New York's AI Companion Models Law
On November 5, 2025, New York Governor Kathy Hochul penned a letter to AI companies operating in the state, announcing that, effective immediately, they would have to implement safety protocols for crisis intervention and clearly notify users that they are interacting with a chatbot and not a real human. Safety protocols under the law include the ability for AI chatbots to detect signs of suicidal ideation and self-harm. It also requires companion models to direct users to “crisis service providers” if they detect these signs.
Unlike California’s “reminder” requirement, which only applies to minors, New York’s AI Companion Models Law mandates that companies provide conspicuous notifications to all users every three hours that they are not interacting with a human. The law, which is enforced by New York’s Attorney General, does not provide an option for individuals to sue, unlike California’s bill.
New Hampshire's HB 143
New Hampshire’s HB 143 was signed into law on August 1, 2025. It criminalized and created the right of an individual to sue if a companion chatbot encouraged, offered, solicited, or recommended that a child engage in acts of self-harm, violence, sexual conduct, or drug or alcohol use. Unlike some other laws, New Hampshire’s act is an extension of the state’s child-endangerment law.
Effective January 1, 2026, the state attorney general can bring an action against the owner or operator of a chatbot company that violates this law. Furthermore, a child, parent, or another party may sue the owner or operator for damages under this law, with damages starting at $1,000 per violation.
Nevada's AB 406
Effective July 1, 2025, Nevada’s AB406 targets chatbots that provide mental or behavioral health services. While it allows individuals or companies that provide mental or behavioral health care services to use AI for administrative purposes, such as scheduling and managing records, it prohibits them from providing or claiming the ability to provide these services.
The bill authorizes the state’s Division of Public and Behavioral Health to investigate potential violations with civil penalties ranging up to $15,000 per violation.
Illinois's WOPR Act HB 1806
On August 4, 2025, Illinois became the first state in the nation to enact a law regulating the use of AI in professional mental health services. The Wellness and Oversight for Psychological Resources Act (WOPR) prohibits AI platforms from providing mental health services or making therapeutic decisions. While mental health providers can continue to use AI for administrative purposes, such as scheduling, it prevents them from making independent treatment decisions, from having a chatbot directly interact with a patient, or from relying on it to detect emotions.
Under the new law, the Illinois Department of Financial and Professional Regulation has the authority to investigate suspected violations and issue fines of up to $10,000.
Maine's Chatbot Disclosure Act
Effective September 24, 2025, Maine’s Chatbot Disclosure Act requires businesses to notify consumers in a “clear and conspicuous manner” that they are not engaging with a real human being. Violations are enforceable under the state’s Unfair Trade Practices Act, which gives individuals an avenue to take legal action against companies that fail to comply with the law.
Utah's Mental Health Chatbot Laws HB 452 & HB 438
Utah has taken several steps to protect consumers from potentially harmful AI. On May 7, 2025, HB 452 took effect, regulating AI-powered mental health chatbots. The law establishes protections for consumers, prohibits certain uses of personal information, requires certain disclosures, and provides enforcement authority to the state’s Division of Consumer Protection.
In addition to HB 452, the state has also moved to pass the Companion Chatbot Safety Act (HB 438), which would regulate these AI providers. Once passed, it would require companion chatbot operators to comply with the Utah Consumer Privacy Act, establish disclosure and data protection requirements, implement additional safety requirements regarding minors, authorize rulemaking on “age assurance standards,” and mandate annual reporting.
Texas's Responsible Artificial Intelligence Governance Act HB 149
The Texas Responsible Artificial Intelligence Governance Act (HB 149) is a broad law covering many aspects of AI, not just chatbots. It establishes the Texas Artificial Intelligence Council, which is responsible for ensuring AI platforms used in the state are ethical. Within the legislation are also safe harbor provisions and the establishment of a regulatory sandbox, which allows AI companies to operate for a short period of time with certain protections.
While the legislation does not have specific regulations for chatbot or companion AI, it does prohibit the use or development of an AI platform that would push a person towards violence, self-harm, or illegal activities. Like some other states, the law is enforced by the Attorney General, but it does not provide a path for individuals to sue directly.
Iowa's SF 2417
Taking effect July 1, 2026, Iowa’s SF 2417 sets forth guidelines and requirements for conversational AI services. It also provides civil penalties for companies that violate its provisions. The law requires conversational AI providers to clearly and conspicuously disclose to minor users that they are speaking to a bot and not a real human. It requires reasonable measures to prevent the production of sexually explicit content or stating that the minor should engage in sexually explicit conduct. Finally, it requires the establishment of protocols for prompts related to self-harm or suicidal ideation, including sending information about a crisis hotline and other resources.
While the bill took effect on July 1, 2026, it does not apply to operators until July 1, 2027. Furthermore, it is enforced by the state Attorney General but does not provide an individual with a direct path to sue.
Tennessee's SB 1580
Tennessee’s SB 1580 took effect on July 1, 2026. The AI legislation prohibits an owner or operator from advertising or representing to the public that the platform is qualified as a mental health professional. Non-compliance with the law is a violation under the Tennessee Consumer Protection Act of 1977 and may constitute an unfair or deceptive practice. Individuals harmed by a violation of the act may take legal action against the company. Violations may also incur civil penalties of $5,000 per violation.
Connecticut’s Public Act 26-15
In an effort to protect consumers, Connecticut lawmakers passed Public Act 26-15. The Connecticut AI bill requires AI platforms to develop ways to provide resources if a prompt brings up self-harm or suicidal ideation. It further regulates AI offering mental health services to minors, encouraging them to engage in self-harm, violence, or illegal activities. Because of the strictness of many of the requirements in Connecticut’s AI legislation, some companies may need to make modifications before the remaining provisions take effect in 2027. Enforcement of the law is under the Attorney General.
Oregon's SB 1546
Oregon’s SB 1546 requires operators of AI companions and AI companion platforms to provide notice to consumers that they are interacting with an artificial agent if a reasonable person would believe they were interacting with a natural person.
The law, which takes effect January 1, 2027, also requires operators of these platforms to develop protocols for prompts related to self-harm or suicidal ideation, and to provide resources and referrals to crisis hotlines if detected.
Finally, Oregon’s AI legislation provides certain protections for children. Unlike some other states, this law allows a person harmed by an AI companion to take legal action against the operator and collect statutory damages of $1,000 per violation.
Washington's Chatbot Disclosure Act HB 2225
Washington’s Chatbot Disclosure Act (HB 2225) takes effect January 1, 2027. The AI legislation regulates companion chatbots by requiring clear and ongoing disclosures that the user is speaking to a bot rather than a human, establishing safeguards to detect and appropriately respond to prompts for self-harm or suicidal ideation, and mandating additional protections for minors, such as restrictions on sexually explicit content. The law leaves enforcement to the Attorney General and does not provide specific provisions that allow a person to sue on their own.
Idaho's Conversational AI Safety Act S 1297
Signed into law on March 31, 2026, Idaho’s Conversational AI Safety Act (S 1297) requires operators to clearly and conspicuously disclose that the user is interacting with artificial intelligence if they could reasonably believe it was a human.
The law, which takes effect on July 1, 2027, also requires operators to adopt protocols for responding to prompts about self-harm or suicidal ideation, including directing them to the proper resources. It prohibits AI services from representing that they can provide mental health services.
Finally, it establishes additional protections for minors, including prohibiting providing rewards for continued or increased engagement. Enforcement is through the Attorney General, and there is no stated path for a person to sue on their own.
Nebraska's Conversational AI Safety Act LB 525
On July 1, 2027, Nebraska’s Conversational AI Safety Act (LB 525) takes effect. The AI legislation requires operators to clearly and conspicuously identify to a minor user that it is interacting with artificial intelligence, not produce sexually explicit content, generate statements that encourage sexually explicit conduct, or make statements that simulate emotional dependence.
The bill also sets forth requirements for the development of protocols when prompts deal with self-harm, such as the referral to a suicide hotline or crisis service. Disclosure requirements apply to all users if a reasonable person would believe they were interacting with a real human. Enforcement of the law is placed with the Attorney General, and there is no path established for a person to sue on their own.
Georgia's AI Chatbot Disclosure and Child Safety Act SB 540
In May 2026, Georgia enacted sweeping AI legislation by passing the AI Chatbot Disclosure and Child Safety Act. Taking effect July 1, 2026, Georgia’s AI bill requires certain disclosures from conversational AI services, mandates protections and safeguards for minor users, requires specific privacy tools, and dictates that these services must adopt a protocol for responding to prompts about suicidal ideation or self-harm.
It is important to note that the bill does not exempt chatbots built into larger platforms, meaning that major tech companies are subject to compliance. The law is enforced by the Attorney General, who may impose civil penalties of up to $10,000 for a knowing violation.
What These Laws Mean for Families
These new laws enacted in dozens of states throughout the country help to provide protections for children, teens, and everyday users of these platforms. Across the board, most AI legislation requires owners or operators to provide clear and conspicuous notice to users that they are interacting with a bot and not an actual human. Additionally, most laws require the platforms to adopt protocols for responding to prompts dealing with self-harm or suicidal ideation. When it comes to minors, many states have additional protections such as prohibiting the dissemination of sexually explicit material or reminding them to take breaks regularly.
Because the industry continues to evolve at a nearly breakneck pace, many states have only recently enacted consumer protection laws. Some laws will not even go into effect until the middle of 2027. Enforcement is still in its infancy, meaning that families may not see action taken immediately after a violation. Even if a company meets certain requirements, it does not mean that someone cannot be harmed. Compliance with disclosure requirements, for instance, does not settle whether a person was harmed by the platform. Families affected by AI companion chatbots are encouraged to follow current chatbot litigation and seek counsel from an experienced attorney if they believe the AI platform caused or contributed to their harm.
Get Answers About an AI Chatbot Injury Today
Wisner Baum continues to investigate and accept AI chatbot injury claims nationwide. These new laws highlight the growing concerns among regulators and consumers about the risks associated with the use of the platforms. If you or a loved one has experienced harm after using an AI chatbot, contact Wisner Baum at (310) 207-3233 for a free, confidential consultation.
