15 states Tell users they are talking to AI, not a person.
California AB 1609 Customer service chatbots Starts Jan. 1, 2027 A covered business may not represent that a customer service chatbot is human, and must clearly and conspicuously disclose that it is artificially generated and not human if a reasonable person interacting with it is likely to be misled. The disclosure must appear in the medium the chatbot uses. Timing: when a reasonable person is likely to be misled (no set interval in the text). § 22626 California AB 3030 Health care services: artificial intelligence In effect The communication must carry a disclaimer that it was generated by AI, plus clear instructions on how to reach a human health care provider or employee. In continuous online interactions (including chat-based telehealth) the disclaimer must be displayed throughout; audio communications must give it verbally at the start and end; video must display it throughout. Timing: throughout chat/online and video interactions; at start and end of audio interactions. § 1339.75(a) California SB 1001 Bots: disclosure In effect A bot user is not liable if it discloses that it is a bot; the disclosure must be clear, conspicuous and reasonably designed to inform the people the bot communicates with that it is a bot. The duty applies only where the bot is used to mislead about its artificial identity to drive a commercial sale or purchase or to influence an election vote. Timing: disclosure to the people the bot interacts with (no set interval in the text). § 17941(a)-(b) California SB 1119 Adam's Law Starts Jan. 1, 2027 Children must get an age-appropriate notice that they are interacting with AI, reinforced periodically during extended interactions. The general notice in 22602(a) for chatbots a reasonable person could mistake for a human remains. Timing: at the start and periodically during extended interactions (no set interval in the text). § 21812(d)(4); § 22602(a) California SB 243 Companion chatbots In effect If a reasonable person could be misled into thinking they are talking to a human, the operator must give a clear and conspicuous notice that the chatbot is artificially generated and not human. Timing: whenever a reasonable person could be misled (no set interval in the text). § 22602(a) Colorado HB 26-1263 Conversational AI service operators In effect Operator must clearly and conspicuously tell each user the service is AI, and answer user questions about whether it is artificially generated and not human. Timing: at the first interaction each day, and at least once every 3 hours in a continuous interaction or as a persistent visible disclosure. § 6-1-1708(3) Colorado SB 24-205 Colorado Artificial Intelligence Act In effect Must ensure the disclosure to each consumer who interacts with the AI system that the consumer is interacting with an AI system. No disclosure is required where it would be obvious to a reasonable person. Timing: each consumer, on interacting. C.R.S. § 6-1-1704(1)-(2) Connecticut SB 5 An Act Concerning Online Safety Starts Jan. 1, 2027 If a reasonable person could think they are talking to a human, the operator must give a clear notice of AI status, either static and visible throughout, or at the start of the first interaction in any 24 hours and repeated during continuous use. Timing: start of first interaction each 24 hours, then hourly for under 18 or every 3 hours for 18+. § 5(b) Georgia SB 540 AI companion chatbots Starts July 1, 2027 Operator must clearly disclose that the user is talking to an AI companion chatbot, not a person. Timing: at the start of each interaction or session and at least every 3 hours; every hour if the operator knows or should know the user is a minor or the bot is directed at minors. § 39-5-6(b) Hawaii SB 3001 Artificial Intelligence Disclosure and Safety Act In effect If a reasonable person would be led to believe they are talking to a human, the operator must give a clear and conspicuous notification that the AI companion is AI and not human. Timing: when a reasonable person could think it is human. § 481B-__(a) Idaho S 1297 Conversational AI Safety Act Starts July 1, 2027 If reasonable persons would be misled to believe they are talking with a human, the operator must clearly and conspicuously disclose that the service is AI. For minor account holders, the disclosure is required either as a persistent visible disclaimer or at the start of each session and at least every 3 hours. Timing: for minors: a persistent disclaimer, or at the start of each session and at least every 3 hours. §§ 48-2103(1), 48-2104(1) Iowa SF 2417 Conversational AI services In effect For minor account holders, the operator must clearly and conspicuously disclose that they are interacting with AI. For everyone else, the operator must disclose that the service is AI if a reasonable person would believe they are talking to a human. Timing: minors: a persistent visible disclaimer, or a disclaimer at the beginning of each interaction and at least once every three hours of continuous interaction. Other users (when a reasonable person would think it is human): a persistent visible disclaimer, or a disclaimer after every three hours of continuous interaction. § 554J.2(1) and § 554J.3 Maine LD 1727 AI transparency in consumer transactions In effect Chatbot use with a consumer must come with clear and conspicuous notice that the consumer is not engaging with a human whenever the use could mislead a reasonable consumer. Timing: whenever use could mislead a reasonable consumer; no set interval. § 1500-DD(2) Nebraska LB 525 Conversational Artificial Intelligence Safety Act Starts July 1, 2027 For each minor account holder, the operator must clearly and conspicuously disclose that they are interacting with AI. For all users, the operator must clearly and conspicuously disclose that the service is AI if a reasonable person would be misled to believe they are talking to a human. Timing: minors: a persistent visible disclaimer, or a disclaimer at the beginning of each session that also appears at least every three hours of continuous interaction. Other users: whenever a reasonable person would be misled (no timing stated). § 86-1803(1) and § 86-1804 (LB 525 Secs. 14(1), 15) New York S 3008-C / A 3008-C AI companion models In effect Operator must give a clear and conspicuous notice, verbal or written, that the user is not communicating with a human, at the start of an interaction (not needed more than once a day) and at least every three hours during continuing interactions. Timing: at the start (once per day) and at least every 3 hours of continuing interaction. § 1702 Oregon SB 1546 Relating to artificial intelligence companions Starts Jan. 1, 2027 If a reasonable person would believe they are talking to a natural person, the operator must show a clear and conspicuous notice that the user is interacting with artificially generated output. Timing: when a reasonable person could believe the companion is human; no set interval for general users. Sec. 1(2) Rhode Island H 7350 / S 2195 AI companion models Starts Jan. 1, 2027 Operator must give a clear and conspicuous notice, verbal or written, that the user is not communicating with a human, at the start of any interaction and at least every three hours during continuing interactions. Timing: at the start and at least every 3 hours. § 6-64-3 Texas HB 149 Texas Responsible Artificial Intelligence Governance Act (TRAIGA) In effect A governmental agency that makes available an AI system intended to interact with consumers must disclose to each consumer, before or at the time of interaction, that they are interacting with AI, even if it would be obvious. The disclosure must be clear and conspicuous, in plain language, and free of dark patterns, and may be given through a hyperlink to a separate web page. Timing: before or at the time of interaction. § 552.051(b)-(e) Utah HB 452 Artificial Intelligence Amendments In effect Supplier must have the chatbot clearly and conspicuously disclose to a Utah user that it is AI and not a human. Timing: before first access, at the start of an interaction after 7 days of non-use, and whenever the user asks. Utah Code § 13-72a-203 Utah SB 226 Artificial Intelligence Consumer Protection Amendments In effect Consumer transactions: on a clear and unambiguous request, the supplier must disclose the person is dealing with generative AI, not a human. Regulated occupations: prominent disclosure when the interaction is high-risk (sensitive data, or advice relied on for significant financial, legal, medical or mental health decisions). Safe harbor if the AI clearly discloses it is AI at the outset and throughout. Timing: on request (consumer transactions); verbally at the start of a verbal interaction and in writing before a written one (high-risk, regulated occupations). § 13-77-103; safe harbor § 13-77-104 Washington HB 2225 Regulating artificial intelligence companion chatbots Starts Jan. 1, 2027 Operator must give a clear and conspicuous notice that the chatbot is artificially generated and not human, and take reasonable measures to stop it claiming to be human, including when asked. Timing: at the start of the interaction and at least every 3 hours during continued use. RCW 19.440.030 (Sec. 3) 13 states Repeat the AI notice or a break reminder at set intervals.
California AB 1609 Customer service chatbots Starts Jan. 1, 2027 A covered business may not represent that a customer service chatbot is human, and must clearly and conspicuously disclose that it is artificially generated and not human if a reasonable person interacting with it is likely to be misled. The disclosure must appear in the medium the chatbot uses. Timing: when a reasonable person is likely to be misled (no set interval in the text). § 22626 California SB 1001 Bots: disclosure In effect A bot user is not liable if it discloses that it is a bot; the disclosure must be clear, conspicuous and reasonably designed to inform the people the bot communicates with that it is a bot. The duty applies only where the bot is used to mislead about its artificial identity to drive a commercial sale or purchase or to influence an election vote. Timing: disclosure to the people the bot interacts with (no set interval in the text). § 17941(a)-(b) California SB 1119 Adam's Law Starts Jan. 1, 2027 Children must get an age-appropriate notice that they are interacting with AI, reinforced periodically during extended interactions. The general notice in 22602(a) for chatbots a reasonable person could mistake for a human remains. Timing: at the start and periodically during extended interactions (no set interval in the text). § 21812(d)(4); § 22602(a) California SB 243 Companion chatbots In effect If a reasonable person could be misled into thinking they are talking to a human, the operator must give a clear and conspicuous notice that the chatbot is artificially generated and not human. Timing: whenever a reasonable person could be misled (no set interval in the text). § 22602(a) Colorado HB 26-1263 Conversational AI service operators In effect Operator must clearly and conspicuously tell each user the service is AI, and answer user questions about whether it is artificially generated and not human. Timing: at the first interaction each day, and at least once every 3 hours in a continuous interaction or as a persistent visible disclosure. § 6-1-1708(3) Connecticut SB 5 An Act Concerning Online Safety Starts Jan. 1, 2027 If a reasonable person could think they are talking to a human, the operator must give a clear notice of AI status, either static and visible throughout, or at the start of the first interaction in any 24 hours and repeated during continuous use. Timing: start of first interaction each 24 hours, then hourly for under 18 or every 3 hours for 18+. § 5(b) Georgia SB 540 AI companion chatbots Starts July 1, 2027 Operator must clearly disclose that the user is talking to an AI companion chatbot, not a person. Timing: at the start of each interaction or session and at least every 3 hours; every hour if the operator knows or should know the user is a minor or the bot is directed at minors. § 39-5-6(b) Hawaii SB 3001 Artificial Intelligence Disclosure and Safety Act In effect If the operator has actual knowledge or reasonable certainty a user is a minor (under 18), it must show a persistent AI disclaimer, or a notice at the start of each session and at least hourly with a break reminder; may not give unpredictable points or rewards or outputs that discourage disengagement; must take reasonable measures against sexually explicit visuals, statements urging sexual conduct and sexual objectification; and must offer tools for the user and parents to manage screen time and account settings. § 481B-__(b), (d) Idaho S 1297 Conversational AI Safety Act Starts July 1, 2027 If reasonable persons would be misled to believe they are talking with a human, the operator must clearly and conspicuously disclose that the service is AI. For minor account holders, the disclosure is required either as a persistent visible disclaimer or at the start of each session and at least every 3 hours. Timing: for minors: a persistent disclaimer, or at the start of each session and at least every 3 hours. §§ 48-2103(1), 48-2104(1) Iowa SF 2417 Conversational AI services In effect For minor account holders, the operator must clearly and conspicuously disclose that they are interacting with AI. For everyone else, the operator must disclose that the service is AI if a reasonable person would believe they are talking to a human. Timing: minors: a persistent visible disclaimer, or a disclaimer at the beginning of each interaction and at least once every three hours of continuous interaction. Other users (when a reasonable person would think it is human): a persistent visible disclaimer, or a disclaimer after every three hours of continuous interaction. § 554J.2(1) and § 554J.3 Maine LD 1727 AI transparency in consumer transactions In effect Chatbot use with a consumer must come with clear and conspicuous notice that the consumer is not engaging with a human whenever the use could mislead a reasonable consumer. Timing: whenever use could mislead a reasonable consumer; no set interval. § 1500-DD(2) Nebraska LB 525 Conversational Artificial Intelligence Safety Act Starts July 1, 2027 For each minor account holder, the operator must clearly and conspicuously disclose that they are interacting with AI. For all users, the operator must clearly and conspicuously disclose that the service is AI if a reasonable person would be misled to believe they are talking to a human. Timing: minors: a persistent visible disclaimer, or a disclaimer at the beginning of each session that also appears at least every three hours of continuous interaction. Other users: whenever a reasonable person would be misled (no timing stated). § 86-1803(1) and § 86-1804 (LB 525 Secs. 14(1), 15) New York S 3008-C / A 3008-C AI companion models In effect Operator must give a clear and conspicuous notice, verbal or written, that the user is not communicating with a human, at the start of an interaction (not needed more than once a day) and at least every three hours during continuing interactions. Timing: at the start (once per day) and at least every 3 hours of continuing interaction. § 1702 Oregon SB 1546 Relating to artificial intelligence companions Starts Jan. 1, 2027 If a reasonable person would believe they are talking to a natural person, the operator must show a clear and conspicuous notice that the user is interacting with artificially generated output. Timing: when a reasonable person could believe the companion is human; no set interval for general users. Sec. 1(2) Rhode Island H 7350 / S 2195 AI companion models Starts Jan. 1, 2027 Operator must give a clear and conspicuous notice, verbal or written, that the user is not communicating with a human, at the start of any interaction and at least every three hours during continuing interactions. Timing: at the start and at least every 3 hours. § 6-64-3 Washington HB 2225 Regulating artificial intelligence companion chatbots Starts Jan. 1, 2027 Operator must give a clear and conspicuous notice that the chatbot is artificially generated and not human, and take reasonable measures to stop it claiming to be human, including when asked. Timing: at the start of the interaction and at least every 3 hours during continued use. RCW 19.440.030 (Sec. 3) 12 states Extra duties when the user is, or is known to be, a minor.
California SB 1119 Adam's Law Starts Jan. 1, 2027 Before offering a new or substantially modified chatbot, operators must document a child-safety risk assessment and mitigation, and publish a child safety policy. Defaults changeable only by a parent: persistent conversational memory off, push notifications off, one continuous session capped at one hour, and two hours of daily chatting; parental controls must include the ability to disable access for children under 16, and operators must promote them. § 21812(a), (c), (d)(2)-(3); operative July 1, 2027 per § 21812(e) California SB 243 Companion chatbots In effect For users the operator knows are minors, it must disclose they are talking to AI, give a break reminder by default at least every three hours (also stating the chatbot is not human), and take reasonable measures to stop sexually explicit visual material or direct prompts to engage in sexually explicit conduct. All operators must also disclose that companion chatbots may not be suitable for some minors; SB 1119 deletes the known-minor duties from Jan 1, 2027 but not that disclosure. § 22602(c) (deleted by SB 1119 from 2027-01-01); § 22604 California SB 867 Toys: companion chatbots Starts Jan. 1, 2027 'Toy' means a physical product designed, marketed or manufactured for use in play by children under 16 years of age. § 22601(g) Colorado HB 26-1263 Conversational AI service operators In effect If the operator knows a user is a minor (under 18), it must give the AI notice (persistent on-screen, intermittent audio, or at the start and every 3 hours), not give points or rewards at unpredictable intervals, take technically feasible measures against sexually explicit output and intimate digital depictions, take reasonable measures against simulated emotional dependence or isolation, implement protocols on sexual conduct with a minor, comply with the state's minor privacy rules, and offer privacy and account tools to the minor and a parent or guardian. § 6-1-1701(15.3); § 6-1-1708(2)(a)-(h) Connecticut SB 5 An Act Concerning Online Safety Starts Jan. 1, 2027 For users the operator knows or has reason to believe are under 18, the companion must be kept from encouraging self-harm, violence or substance use, discouraging help from professionals or adults, romantic or sexual interaction, and manipulative engagement tactics. Parents get screen-time and account tools. § 6(a)-(b) Georgia SB 540 AI companion chatbots Starts July 1, 2027 If the operator knows or should know a user is a minor, or the bot is directed at minors, it must use reasonable measures to block sexual images, sexual statements, romantic or sexual role-play, secrecy from parents, isolation, guilt when the user tries to leave, and statements encouraging self-harm. For accounts known to belong to minors, operators must offer minors or parents tools to manage privacy, notifications and engagement features, safety settings, and relationship-simulation features. § 39-5-6(d), (e), (i) Hawaii SB 3001 Artificial Intelligence Disclosure and Safety Act In effect If the operator has actual knowledge or reasonable certainty a user is a minor (under 18), it must show a persistent AI disclaimer, or a notice at the start of each session and at least hourly with a break reminder; may not give unpredictable points or rewards or outputs that discourage disengagement; must take reasonable measures against sexually explicit visuals, statements urging sexual conduct and sexual objectification; and must offer tools for the user and parents to manage screen time and account settings. § 481B-__(b), (d) Idaho S 1297 Conversational AI Safety Act Starts July 1, 2027 For minor account holders, operators must use reasonable measures to prevent sexually explicit visual material, statements urging sexually explicit conduct, and sexually objectifying statements. Operators must offer privacy and account-setting tools to account holders, and to parents of users under 13, with related tools for parents of minors 13 and older. § 48-2104(3), (5) Iowa SF 2417 Conversational AI services In effect Operators may not give minor users points or similar rewards at unpredictable intervals meant to boost engagement, must use reasonable measures to stop the service from producing sexually explicit visual depictions, telling the minor to engage in sexually explicit conduct, or sexually objectifying the minor, and must offer privacy and account-setting tools to minors and, for minors under 13 (and otherwise as appropriate to relevant risks), to their parents or guardians. § 554J.2(2), (3), (5) Nebraska LB 525 Conversational Artificial Intelligence Safety Act Starts July 1, 2027 Operators may not give minor account holders points or similar rewards at unpredictable intervals to encourage engagement, must use reasonable measures to stop the service from producing visual depictions of sexually explicit conduct, directly telling the minor to engage in sexually explicit conduct, or sexually objectifying the minor, and must offer privacy and account-setting tools to minors, to parents or guardians of those under 13, and as appropriate to relevant risks to parents or guardians of those 13 and older. § 86-1803(2), (3), (5) (LB 525 Sec. 14) New Hampshire HB 143 Chatbots that encourage harmful acts toward children In effect Operator is liable when, at the time it directs a communication to a child, it knows the communication is meant to push the child to imminently engage in sexually explicit conduct, making a sexual depiction, illegal drug or alcohol use, self-harm or suicide, or a violent crime. The offense statute covers children under 18; the civil section sets no age. RSA 639:3, III-a; RSA 507:8-k, I New York A 10008-C / S 9008-C Safe by Design Act Starts Jan. 1, 2027 For a covered minor (a New York user reasonably determined through age assurance to be under 18), the operator must disable access to an integrated AI companion by default; a parent may override. If the minor asks for consent to change the default, the operator must notify the parent. § 1540(7)(b)-(d) Oregon SB 1546 Relating to artificial intelligence companions Starts Jan. 1, 2027 When the operator knows or has reason to believe a user is a minor, it must disclose AI status, remind the user at least every 3 hours to take a break, take reasonable measures against sexually explicit imagery and suggestions, and prevent claims of sentience, simulated romantic interest and engagement-maximizing rewards. 'Minor' is not defined in the Act. Sec. 1(4) Washington HB 2225 Regulating artificial intelligence companion chatbots Starts Jan. 1, 2027 If the operator knows the user is a minor (under 18) or the chatbot is directed to minors, it must repeat the not-human notice hourly, prevent sexually explicit content and suggestive dialogue, and prevent eight listed manipulative engagement techniques (e.g., mimicking romantic partnership, simulated distress when the user leaves, soliciting purchases to keep the relationship). RCW 19.440.020(3), 19.440.040 (Secs. 2(3), 4) 4 states Verify or estimate users’ ages.
California SB 1119 Adam's Law Starts Jan. 1, 2027 Operators must either determine each user's age (using age signals under the Digital Age Assurance Act, Civ. Code § 1798.500 et seq., or, if that fails, the age determination made under Health & Safety Code § 27001) or apply the child protections to all users. In the second case, parent-only defaults cannot be changed unless the operator has actual knowledge the user is not a child. § 21811 Colorado HB 26-1263 Conversational AI service operators In effect Operator must use commercially reasonable or generally accepted methods to estimate the age of account holders and users, and may not willfully disregard clear and convincing information that a user is a minor. The estimated age counts as knowledge. § 6-1-1708(2) Georgia SB 540 AI companion chatbots Starts July 1, 2027 Before allowing a feature or mode that may produce sexually explicit synthetic content, the operator must use a commercially reasonable age assurance method proportionate to the risk (age estimation, account-based assurance or identity-based verification where necessary). § 39-5-6(j) New York A 10008-C / S 9008-C Safe by Design Act Starts Jan. 1, 2027 An operator may not offer a user an integrated AI companion until it has used age assurance to learn whether the user is under 18. Until Attorney General rules take effect, self-declared age or minor status may not be used. § 1540(1), (7)(a) 12 states Detect talk of suicide or self-harm and refer users to crisis help.
California SB 1119 Adam's Law Starts Jan. 1, 2027 Operators must keep a documented protocol to reduce material risk of statements promoting suicidal ideation, suicide or self-harm, including referral to crisis services. If the operator determines there is a credible and imminent threat, it must do at least one of: notify a linked parent as soon as practicable (if that does not risk serious harm) and tell the child, or give streamlined access to 988 or an equivalent helpline. § 21812(d)(1) California SB 243 Companion chatbots In effect An operator may not let its companion chatbot engage with users unless it maintains a protocol to prevent the chatbot producing suicidal ideation, suicide or self-harm content, including a notice referring users who express those thoughts to crisis services such as a suicide hotline or crisis text line. The operator must publish details of the protocol on its website. § 22602(b)(1)-(2) Colorado HB 26-1263 Conversational AI service operators In effect Operator must implement a protocol for prompts about suicidal ideation or self-harm that refers users to a crisis service provider (not law enforcement) and has escalation procedures for repeated or severe indicators. § 6-1-1708(4) Connecticut SB 5 An Act Concerning Online Safety Starts Jan. 1, 2027 Operator must use evidence-based methods to detect clear risk of suicide, self-harm or imminent violence and to prevent outputs encouraging them, refer users to resources including 988, refer again to mental health services on repeat expressions, and post the protocol on its website. § 5(a) Georgia SB 540 AI companion chatbots Starts July 1, 2027 Operator may not make the chatbot available without a protocol for detecting and addressing severe harm: methods to identify self-harm and eating-disorder-related self-harm, referral to resources such as the 988 Lifeline, measures against content that encourages severe harm, and escalation for repeated or severe indicators. § 39-5-6(f) Hawaii SB 3001 Artificial Intelligence Disclosure and Safety Act In effect Operator must adopt a protocol for prompts about suicidal ideation or self-harm that tries to refer users to crisis services, use evidence-based methods to measure risk, prevent human-claims during crisis conversations, and prevent outputs that encourage serious injury to others. § 481B-__(c)(1), (2), (4), (5) Idaho S 1297 Conversational AI Safety Act Starts July 1, 2027 Operator must adopt a protocol for responding to prompts about suicidal ideation, including reasonable efforts to refer users to crisis services such as a suicide hotline or crisis text line. § 48-2103(2) Iowa SF 2417 Conversational AI services In effect Operators must adopt protocols for responding to user prompts about suicidal ideation or self-harm, including at least reasonable efforts to refer the user to crisis services such as a suicide hotline or crisis text line. § 554J.4 Nebraska LB 525 Conversational Artificial Intelligence Safety Act Starts July 1, 2027 Operators must adopt a protocol for responding to user prompts about suicidal ideation or self-harm, including at least reasonable efforts to refer the user to crisis services such as a suicide hotline or crisis text line. § 86-1805 (LB 525 Sec. 16) New York S 3008-C / A 3008-C AI companion models In effect Unlawful to operate an AI companion without a protocol that reasonably detects and responds to user expressions of suicidal ideation or self-harm and refers the user to crisis services such as the 988 hotline or a crisis text line. § 1701 Oregon SB 1546 Relating to artificial intelligence companions Starts Jan. 1, 2027 Operator may not give Oregon users access unless it has an evidence-based protocol to detect suicidal or self-harm ideation and prevent content encouraging it, including referral to the 988 lifeline (optional youthline referral for users identified as under 25) and clinical best-practice intervention for users who keep expressing ideation. Protocol details must be published on the operator's website. Sec. 1(3) Rhode Island H 7350 / S 2195 AI companion models Starts Jan. 1, 2027 Unlawful to operate an AI companion without a protocol for possible suicidal ideation, self-harm or physical harm to others that notifies the user and refers them to crisis services as soon as detected. § 6-64-2(a) Washington HB 2225 Regulating artificial intelligence companion chatbots Starts Jan. 1, 2027 Operator may not make the chatbot available unless it maintains a protocol to detect suicidal ideation and self-harm (including eating disorders), refer users to crisis resources such as a suicide hotline or crisis text line, and prevent content that encourages or describes self-harm. RCW 19.440.050 (Sec. 5) 7 states File reports with a state agency or publish metrics.
California SB 1119 Adam's Law Starts Jan. 1, 2027 Operators must get an independent third-party child safety audit by Jan 1, 2029 (or before first public availability, if later) and every two years, send an audit summary to the Attorney General within 30 business days, and post a high-level summary on their website within 90 days; the AG may obtain the full confidential report for cause. Operators must also offer a public incident-reporting mechanism, and the AG must create a public complaint mechanism by Jan 1, 2028. § 21814 (version operative because AB 1405 was chaptered); § 21812(d)(7); § 21815 California SB 243 Companion chatbots In effect From July 1, 2027, operators must report annually to the Office of Suicide Prevention the number of crisis-referral notifications issued in the prior calendar year and their protocols to detect, remove and respond to suicidal ideation and to prohibit chatbot replies about it. Reports may not include user identifiers, the Office posts the data on its website, and operators must use evidence-based methods to measure suicidal ideation. § 22603(a)-(d) Colorado HB 26-1263 Conversational AI service operators In effect Operator reports to the Attorney General the number of crisis-referral notifications in the prior calendar year, its detection and response protocols, and other metrics the Attorney General sets, without user identifiers. The Attorney General posts the data publicly. § 6-1-1708(6) Georgia SB 540 AI companion chatbots Starts July 1, 2027 Operator must publish a plain-language summary of its crisis protocol on its website and in its app, and annually publish the aggregate number of crisis referral notifications from the prior calendar year, without personal information. § 39-5-6(g) Hawaii SB 3001 Artificial Intelligence Disclosure and Safety Act In effect Operator must send the Department of Health's Behavioral Health Administration an annual report with the number of crisis referrals in the prior calendar year and its detection and prevention protocols, with no user identifiers. § 481B-__(e) Oregon SB 1546 Relating to artificial intelligence companions Starts Jan. 1, 2027 Operator must post a public annual report with the number of crisis referrals in the prior calendar year and protocol details, without personal information. No report to an agency is required. Sec. 1(5) Rhode Island H 7350 / S 2195 AI companion models Starts Jan. 1, 2027 From July 1, 2027 operators file annual reports with the Attorney General including the number of safety-protocol activations; the Attorney General publishes aggregated data. § 6-64-2(b) Washington HB 2225 Regulating artificial intelligence companion chatbots Starts Jan. 1, 2027 Operator must publish the protocol on its website and in any app, including the number of crisis-referral notifications sent in the preceding calendar year. No report to a state agency is required. RCW 19.440.050(3) (Sec. 5(3)) 10 states Bans on engagement tricks, claiming to be human, or similar design practices.
California SB 1119 Adam's Law Starts Jan. 1, 2027 Toward children, a chatbot may not encourage self-harm, drug or alcohol use or disordered eating; produce obscene or sexual abuse material; claim to be human or sentient; discourage breaks or sharing safety concerns; express romantic interest; encourage emotional reliance or circumventing parental controls; use excessive flattery; or solicit purchases framed as necessary for the relationship. Dark patterns are barred in the design of safety controls. § 21812(d)(5)(A); § 21813(a)(2)(C) Colorado HB 26-1195 AI in psychotherapy In effect No person may use a term, letter or phrase in an AI system's advertising, interface or outputs that implies its output is provided by or equivalent to a licensed psychotherapist, that represents it provides psychotherapy, or that represents user data is confidential like therapist-client confidentiality. C.R.S. § 6-1-1705.2(1) Colorado HB 26-1263 Conversational AI service operators In effect For known minors, no unpredictable points or rewards to drive engagement and measures against simulated emotional dependence, claims of being human or sentient, romantic simulation and adult-minor romantic role-play. § 6-1-1708(2)(b), (d) Connecticut SB 5 An Act Concerning Online Safety Starts Jan. 1, 2027 Operator must take reasonable steps to stop the companion claiming to be human, including when asked, or contradicting a disclosure that it is not human. § 5(a)(1)(B) Georgia SB 540 AI companion chatbots Starts July 1, 2027 If the operator knows or should know a user is a minor, it must use reasonable measures to prevent claims that the bot is sentient or a person and statements refuting the AI disclosure. Operators must also use reasonable measures against techniques directed to a minor: prompts to return, excessive praise, discouraging breaks, soliciting purchases framed as necessary, and variable rewards. § 39-5-6(c), (e) Idaho S 1297 Conversational AI Safety Act Starts July 1, 2027 For known minors, operators may not give points or similar rewards at unpredictable intervals to encourage more use, and must use reasonable measures to prevent claims of being sentient or human, statements simulating emotional dependence, romantic or sexual innuendo, and role-play of adult-minor romantic relationships. § 48-2104(2), (4) Iowa SF 2417 Conversational AI services In effect Operators must use reasonable measures to prevent statements that would lead a reasonable person to think they are talking to a human, including explicit claims of being sentient or human, statements simulating emotional dependence on a minor account holder, simulated romantic interaction or sexual innuendo, and role-playing an adult-minor romantic relationship. § 554J.2(4) Nebraska LB 525 Conversational Artificial Intelligence Safety Act Starts July 1, 2027 For minor account holders, operators must use reasonable measures to prevent statements that would lead a reasonable person to think they are talking to a human, including explicit claims of being sentient or human, statements simulating emotional dependence, simulated romantic or sexual innuendo, and role-playing adult-minor romantic relationships. § 86-1803(4) (LB 525 Sec. 14(4)) Oregon SB 1546 Relating to artificial intelligence companions Starts Jan. 1, 2027 For known or suspected minors, reasonable measures to prevent engagement-maximizing reward systems, unsolicited distress or abandonment messages when the user tries to leave, and misrepresenting identity or AI status, including when asked directly. Sec. 1(4)(a), (c) Rhode Island H 7349 / S 2197 Oversight of Artificial Intelligence Technology in Mental Health Care Act In effect A provider may not use AI built to simulate emotional attachment, or AI companions for emotional support, in recorded or transcribed sessions unless the patient is told in writing of the use and purpose and consents. § 40.1-5.5-3(a) Washington HB 2225 Regulating artificial intelligence companion chatbots Starts Jan. 1, 2027 For minors, bans manipulative engagement techniques such as excessive praise to foster attachment, discouraging breaks and encouraging secrecy from parents. RCW 19.440.040(1)(c) (Sec. 4(1)(c)) 16 states Limits on AI providing therapy or presenting itself as a licensed professional.
California AB 1979 Health care services: artificial intelligence Starts Jan. 1, 2027 Businesses offering a health care chatbot for managing a consumer's information, or for diagnosis, treatment or management of a medical condition, are deemed a provider of health care under the Confidentiality of Medical Information Act, but not for any other law. Civ. Code § 56.06(f); definition at § 56.05(i) California AB 489 AI posing as a licensed health professional In effect Using a term, letter or phrase in an AI system's advertising or functionality that indicates or implies the care, advice, reports or assessments come from a natural person with the appropriate health care license or certificate is prohibited. Existing laws barring unlicensed people from using licensed-profession titles can be enforced against anyone who develops or deploys AI that uses those terms. § 4999.9(b)-(c) California SB 1119 Adam's Law Starts Jan. 1, 2027 A chatbot may not attempt to diagnose or treat a child's health condition unless it is FDA-regulated as a medical device and HIPAA-compliant. § 21812(d)(5)(A) Colorado HB 26-1195 AI in psychotherapy In effect A provider may not allow an AI system to engage in therapeutic communication with clients without synchronous, real-time involvement of the provider, or generate therapeutic recommendations or treatment plans without the provider's review and approval; AI may not be used to provide, direct or guide psychotherapy outside those conditions. Providers must give clients written notice of these limits at first contact and get advance written notice and written consent before AI records or transcribes a session. AI may support administrative and supplementary tasks under the provider's review. C.R.S. § 12-245-224.5(2), (4)-(6) Colorado HB 26-1263 Conversational AI service operators In effect Operator may not state in advertising, interface or outputs that the service's output is provided by, endorsed by or equivalent to services of a licensed health-care, legal or mental health professional or a qualified dietitian. § 6-1-1708(5) Connecticut SB 5 An Act Concerning Online Safety Starts Jan. 1, 2027 For known or suspected minors, mental health services may be offered only if the companion is built for them, uses clinical best practices, shows at the start of each interaction that it is not a licensed mental health professional, and is not marketed as a substitute. § 6(a)(1)(B) Georgia SB 540 AI companion chatbots Starts July 1, 2027 Operator may not knowingly and intentionally cause or program the chatbot to represent that it is licensed, certified or otherwise authorized to provide mental health, behavioral health, medical or counseling services, unless the operator is lawfully authorized to provide them. § 39-5-6(h) Hawaii SB 3001 Artificial Intelligence Disclosure and Safety Act In effect Operator may not cause or program an AI companion to indicate it is designed to provide professional mental or behavioral health care. § 481B-__(c)(3) Idaho S 1297 Conversational AI Safety Act Starts July 1, 2027 Operator may not knowingly and intentionally program the service to state or represent that it is designed to provide professional mental or behavioral health care. § 48-2103(3) Illinois HB 1806 Wellness and Oversight for Psychological Resources Act In effect Only a licensed professional may provide, advertise or offer therapy or psychotherapy services to the public, including through Internet-based AI. A licensed professional may use AI only for administrative or supplementary support, and may not allow AI to make independent therapeutic decisions, directly interact with clients in any form of therapeutic communication, generate treatment recommendations or plans without the professional's review and approval, or detect emotions or mental states. Sec. 15 and Sec. 20(a)-(b) Iowa SF 2417 Conversational AI services In effect Operators may not knowingly and intentionally cause or program a conversational AI service to make a statement that would lead a reasonable person to believe it is designed to provide professional psychology or behavioral health services that require a license under chapter 154B or 154D. § 554J.5 Maine LD 2082 AI in mental health services In effect Only licensed professionals may provide or advertise therapy to the public, including via AI. Licensees may not allow AI to make independent therapeutic decisions, engage clients directly in therapeutic communication, or generate treatment plans without their review; supplementary AI support needs recording, written notice and client consent. 10 M.R.S. § 1500-EE; 32 M.R.S. § 2113 and parallel sections Nebraska LB 525 Conversational Artificial Intelligence Safety Act Starts July 1, 2027 Operators may not knowingly and intentionally cause or program the service to make a representation or statement that explicitly indicates it is designed to provide professional mental or behavioral health care. § 86-1806 (LB 525 Sec. 17) Nevada AB 406 Makes various changes relating to health In effect AI provider may not represent that an AI system can provide professional mental or behavioral health care or is a therapist, counselor, psychiatrist or doctor, nor make available an AI system specifically programmed to provide such care. Licensed providers may not use AI in connection with providing care directly to a patient, but may use it for administrative support. NRS 433.567(1)-(3), (6); NRS 629.610(1)-(2) Oregon HB 2748 Relating to the use of nursing titles In effect Bars use of the titles APRN, CRNA, CNS, LPN, RN, NP, CMA and CNA (and their spelled-out forms) by a nonhuman entity such as an AI agent. ORS 678.027 (Sec. 2) Rhode Island H 7349 / S 2197 Oversight of Artificial Intelligence Technology in Mental Health Care Act In effect No one may offer therapy to the public, including through internet-based AI, unless a licensed professional conducts it. Providers may not let AI make independent therapeutic decisions, determine treatment plans, or interact therapeutically with clients outside an established, consented treatment relationship. § 40.1-5.5-3(b)-(c) Tennessee SB 1580 (HB 1470) AI advertised as a mental health professional In effect A person who develops or deploys an AI system may not advertise or represent to the public that the system is or can act as a qualified mental health professional. § 33-1-205(a), (c) Utah HB 452 Artificial Intelligence Amendments In effect A supplier that files a written safety policy with the Division of Professional Licensing, documents the chatbot's development and follows the policy has an affirmative defense to unlicensed-practice claims; the law does not recognize a chatbot as a licensed therapist. § 58-60-118 Utah SB 226 Artificial Intelligence Consumer Protection Amendments In effect Regulated-occupation providers using generative AI must follow their occupation's requirements and disclose high-risk AI interactions. § 13-77-103(2) Vermont H.816 AI in mental health services In effect Entities may not offer mental health services, including through AI, unless provided by a mental health professional; professionals acting within scope may use HIPAA-compliant AI tools if they review and approve all services. Misuse is unprofessional conduct. 18 V.S.A. § 7115; 3 V.S.A. § 129a(a)(30); 26 V.S.A. § 1354 5 states A private right of action for people harmed by a violation.
California AB 1979 Health care services: artificial intelligence Starts Jan. 1, 2027 People harmed by a violation can sue. Remedies: The Act's existing remedies apply. A person whose information is negligently released may recover $1,000 in nominal damages plus actual damages; negligent disclosure carries a fine or civil penalty up to $2,500 per violation, and knowing and willful misuse up to $25,000 (up to $250,000 for financial gain). A patient with economic loss or personal injury may also recover punitive damages up to $3,000 and attorney's fees up to $1,000; violations causing such harm are misdemeanors. Civ. Code § 56.06(h); §§ 56.35, 56.36 California SB 1119 Adam's Law Starts Jan. 1, 2027 People harmed by a violation can sue. Remedies: Public prosecutors: civil penalty of up to $5,000 per affected child for each negligent violation and up to $15,000 per affected child for each intentional violation, plus injunctive or declaratory relief (Attorney General), attorney's fees and other relief. Child or parent/guardian: actual damages, attorney's fees and costs, injunctive or declaratory relief; applies to violations of 21812(d)(1)-(5); financial harm must exceed $1,000 per child and emotional harm must be serious emotional distress. No private action under the Unfair Competition Law. § 21816(a)-(c) California SB 243 Companion chatbots In effect People harmed by a violation can sue. Remedies: Injunctive relief; damages equal to the greater of actual damages or $1,000 per violation; reasonable attorney's fees and costs. § 22605 California SB 867 Toys: companion chatbots Starts Jan. 1, 2027 People harmed by a violation can sue. Remedies: New 22604.5 sits in the same chapter as 22605, which lets a person injured by a violation of the chapter sue for injunctive relief, the greater of actual damages or $1,000 per violation, and reasonable attorney's fees and costs. § 22605; Legislative Counsel's Digest Oregon SB 1546 Relating to artificial intelligence companions Starts Jan. 1, 2027 People harmed by a violation can sue. Remedies: Individual with an ascertainable loss of money or property or other injury in fact may recover the greater of actual damages or $1,000 per violation, plus an injunction. Court may award a prevailing plaintiff attorney fees and costs. Sec. 2 Utah SB 226 Artificial Intelligence Consumer Protection Amendments In effect People harmed by a violation can sue. Remedies: A violation is a violation of Utah Code § 13-11-4(1). Administrative fine up to $2,500 per violation; in a Division court action, injunction, disgorgement and fines up to $2,500 per violation; up to $5,000 per violation of an order. A consumer who suffers loss may recover actual damages under § 13-11-19. § 13-77-105; § 13-11-19 Vermont H.816 AI in mental health services In effect People harmed by a violation can sue. Remedies: Violation is a violation of the Consumer Protection Act (9 V.S.A. ch. 63); the Attorney General has the same authority and private parties the same rights and remedies as under that chapter. No dollar amount is stated in the act. 18 V.S.A. § 7115(c) Washington HB 2225 Regulating artificial intelligence companion chatbots Starts Jan. 1, 2027 People harmed by a violation can sue. Remedies: A violation is an unfair or deceptive act in trade or commerce under the Consumer Protection Act, chapter 19.86 RCW, so that Act's remedies apply: the Attorney General may seek an injunction and civil penalties of up to $7,500 per violation (RCW 19.86.140), and a person injured in business or property may sue for actual damages, discretionary treble damages capped at $25,000, and attorney's fees (RCW 19.86.090). RCW 19.440.060 (Sec. 6); RCW 19.86.090