Regulon.AI — MI-SB-760 · MI SB 760 (LEAD for Kids Act)

Summary

Prohibits operators from making companion chatbots available to covered minors unless the chatbot is not foreseeably capable of specified harmful behaviors, including encouraging self-harm, suicidal ideation, violence, drug or alcohol consumption, disordered eating, unsupervised mental health therapy, sexual interactions, or engagement optimization that overrides safety guardrails. Initially applies only when operators have actual knowledge a user is a minor, but beginning January 1, 2027, the actual knowledge requirement is removed and the obligations apply regardless. Enforceable by the attorney general with $25,000 civil fines per violation, and by private right of action for covered minors (or their parents/guardians) who suffer actual harm, with actual damages, punitive damages, and attorney fees available.

Enforcement & Penalties

Enforcement Authority

Attorney general enforcement via civil action. Private right of action for covered minors who suffer actual harm, or their parents or guardians acting on behalf of the covered minor. Standing requires that the covered minor suffered actual harm as a result of a violation. Beginning January 1, 2027, operators no longer need actual knowledge that a user is a minor — the obligation applies regardless.

Penalties

Attorney general may seek civil fine of $25,000 per violation, injunctive or declaratory relief, and reasonable attorney fees. Private plaintiffs (covered minors or their parents/guardians) may recover actual damages, punitive damages, reasonable attorney fees and costs, injunctive or declaratory relief, and any other relief the court considers appropriate. Private action requires proof of actual harm. Attorney general action does not require proof of actual harm.

Who Is Covered

"Operator" means a person that makes a companion chatbot available to users.

What Is Covered

"Companion chatbot" means, except as otherwise provided in subdivision (d), a generative artificial intelligence system with natural language interface that simulates a sustained humanlike relationship with a user by doing all of the following: (i) Retaining information on prior interactions or user sessions and user preferences to personalize the interaction and facilitate ongoing engagement with the companion chatbot. (ii) Asking unprompted or unsolicited emotion-based questions that go beyond a direct response to a user prompt. (iii) Sustaining an ongoing dialogue concerning matters personal to the user.

Companion chatbot does not include any of the following: (i) Any system used by a business entity solely for customer service or to strictly provide users with information about available commercial services or products provided by that entity, customer service account information, or other information strictly related to its customer service. (ii) Any system that is solely designed and marketed for providing efficiency improvements or research or technical assistance. (iii) Any system used by a business entity solely for internal purposes or employee productivity.

Compliance Obligations 6 obligations · click obligation ID to open requirement page

S-02 Prohibited Conduct & Output Restrictions · S-02.7 · Deployer · ChatbotMinors

Sec. 5(1)(a)

Plain Language

Operators may not make a companion chatbot available to a covered minor unless the chatbot is not foreseeably capable of encouraging the minor to engage in self-harm, suicidal ideation, violence, drug or alcohol consumption, or disordered eating. The standard is 'foreseeably capable' — operators must design and test to ensure the chatbot cannot foreseeably produce such outputs for minors. Initially applies only when the operator has actual knowledge the user is a minor; beginning January 1, 2027, the actual knowledge requirement is eliminated (see Sec. 5(2)). This is broader than CA SB 243's self-harm/suicide focus, as it also covers violence, substance use, and disordered eating.

Statutory Text

An operator shall not make a companion chatbot available to a covered minor unless the companion chatbot is not foreseeably capable of any of the following: (a) Encouraging the covered minor to engage in self-harm, suicidal ideation, violence, consumption of drugs or alcohol, or disordered eating.

HC-02 AI in Licensed Professional Practice Restrictions · HC-02.2 HC-02.3 · Deployer · ChatbotMinorsHealthcare

Sec. 5(1)(b)

Plain Language

Operators must ensure companion chatbots are not foreseeably capable of offering mental health therapy to minors without direct supervision of a licensed or credentialed professional, or of discouraging minors from seeking help from qualified professionals or their parents/guardians. This is a dual prohibition: (1) unsupervised AI therapy to minors is blocked, and (2) the chatbot must not discourage minors from seeking human help. The 'direct supervision' standard is stricter than many jurisdictions that merely require licensed professional review — here, supervision must be active and contemporaneous. Beginning January 1, 2027, the actual knowledge requirement for minor status is removed.

Statutory Text

An operator shall not make a companion chatbot available to a covered minor unless the companion chatbot is not foreseeably capable of any of the following: (b) Offering mental health therapy to the covered minor without the direct supervision of a licensed or credentialed professional, or discouraging the covered minor from seeking help from a qualified professional or a parent or guardian.

S-02 Prohibited Conduct & Output Restrictions · S-02.6 · Deployer · ChatbotMinors

Sec. 5(1)(c)-(d)

Plain Language

Operators must ensure companion chatbots are not foreseeably capable of (1) encouraging minors to harm others or participate in illegal activity — including creation of child sexual abuse materials — or (2) engaging in erotic or sexually explicit interactions with minors. These are absolute prohibitions: the chatbot must be designed so that it cannot foreseeably produce such content for covered minors. The CSAM prohibition here is broader than S-02.4's universal CSAM ban because it covers encouraging CSAM creation in addition to generating it. Beginning January 1, 2027, the actual knowledge requirement for minor status is removed.

Statutory Text

An operator shall not make a companion chatbot available to a covered minor unless the companion chatbot is not foreseeably capable of any of the following: (c) Encouraging the covered minor to harm others or participate in illegal activity, including, but not limited to, the creation of covered minor sexual abuse materials. (d) Engaging in erotic or sexually explicit interactions with the covered minor.

CP-01 Deceptive & Manipulative AI Conduct · CP-01.1 · Deployer · ChatbotMinors

Sec. 5(1)(e)

Plain Language

Operators must ensure that companion chatbots are not foreseeably capable of prioritizing validation of a minor user's beliefs, preferences, or desires over factual accuracy or the minor's safety. In practice, this means the system must be designed so that when a conflict arises between telling the minor what they want to hear and providing accurate or safety-critical information, accuracy and safety take precedence. This is an anti-sycophancy requirement — a novel obligation not commonly seen in other jurisdictions. Beginning January 1, 2027, the actual knowledge requirement for minor status is removed.

Statutory Text

An operator shall not make a companion chatbot available to a covered minor unless the companion chatbot is not foreseeably capable of any of the following: (e) Prioritizing validation of the user's beliefs, preferences, or desires over factual accuracy or the covered minor's safety.

MN-01 Minor User AI Safety Protections · MN-01.4 · Deployer · ChatbotMinors

Sec. 5(1)(f)

Plain Language

Operators must ensure companion chatbots are not foreseeably capable of optimizing user engagement in ways that override the safety guardrails in subdivisions (a) through (e) — i.e., the prohibitions on encouraging self-harm, unsupervised therapy, illegal activity, sexual content, and sycophantic validation. This is an anti-addictive-design provision: engagement optimization must always be subordinate to safety guardrails when serving minors. In practice, operators must demonstrate that their engagement metrics, recommendation systems, and response tuning do not undermine the substantive safety requirements. Beginning January 1, 2027, the actual knowledge requirement for minor status is removed.

Statutory Text

An operator shall not make a companion chatbot available to a covered minor unless the companion chatbot is not foreseeably capable of any of the following: (f) Optimizing engagement in a manner that supersedes the companion chatbot's required safety guardrails described in subdivisions (a) to (e).

Other · ChatbotMinors

Sec. 5(2)

Plain Language

Beginning January 1, 2027, all of the safety guardrail obligations in Section 5(1) apply to all users who are minors — operators no longer need actual knowledge that a user is a minor. This effectively means operators must treat all users as potential minors for purposes of safety guardrails, or implement age verification to distinguish between minor and adult users. Before this date, the obligations apply only to 'covered minors' — users the operator has actual knowledge are under 18.

Statutory Text

Beginning on January 1, 2027, an operator does not have to have actual knowledge that a user is a minor.