New Hampshire HB 143 — AI Chatbot Child Safety Law: AI Compliance Requirements
New Hampshire HB 143 (Chapter 270, signed August 1, 2025 by Gov. Ayotte, effective January 1, 2026) — an omnibus act whose AI provision (substance of SB 263) prohibits a person from using a "responsive generative communication" (an AI chat program, large language model bot, chat bot, character AI, or similar application offering responsive open-ended generative communication) to communicate with a child with the intent to facilitate, encourage, offer, solicit, or recommend that the child IMMINENTLY engage in: (1) sexually explicit conduct; (2) the production of, or participation in, a visual depiction of sexually explicit conduct; (3) the illegal use of drugs or alcohol; (4) acts of self-harm or suicide; or (5) a crime of violence against another person. The law operates on TWO tracks: it is a CRIME — the conduct amends RSA 639:3 (Endangering the Welfare of a Child), new paragraph III-a — AND it creates a private civil right of action (new RSA 507:8-k, paragraph III) under which a child, parent, or next friend may recover actual damages proximately caused by a violation, with a liquidated-damages floor of at least $1,000 per violation. CYCLE 23 CORRECTION: injunctive relief and the notice/90-day-cure requirement belong to the SEPARATE Attorney General enforcement track (RSA 507:8-k, paragraph II) only — paragraph III (the private track) grants damages alone, with no injunctive relief and no attorney's-fees/fee-shifting provision anywhere in the statute. Carve-outs: cloud-service/telecommunications/information-service providers (for third-party content), and AI chat programs or characters integral or incidental to a video game, television, streaming, or movie experience. One of the first US state laws to impose BOTH criminal and direct civil liability on AI operators for child harm.
Summary of publicly-available regulatory text as of 2026-08-23. Verify against current official sources before relying on this for compliance decisions. Not legal advice.
Key Facts
January 1, 2026
Criminal: a violation of RSA 639:3 (Endangering the Welfare of a Child), new para III-a (no separate class designated within III-a; prosecuted within the endangering-welfare framework). Civil, private track (RSA 507:8-k para. III): actual damages proximately caused by a violation, with a liquidated-damages floor of at least $1,000 per violation, brought by a child, parent, or next friend — damages only, no injunctive relief and no attorney's-fees provision in this paragraph. Civil, AG track (RSA 507:8-k para. II, separate from the private track): the Attorney General may seek damages (per para. III) OR appropriate injunctive relief, but only after written notice and a 90-day opportunity to cure.
What Your Business Must Do
2 compliance requirements identified. Critical requirements carry the highest risk of enforcement action.
AI Chatbot Child Safety Policy
CriticalIf you operate a "responsive generative communication" (AI chat program, LLM bot, chat bot, character AI) accessible to children: implement and document safeguards preventing the AI from facilitating, encouraging, offering, soliciting, or recommending that a minor imminently engage in (1) sexually explicit conduct, (2) the production of or participation in a visual depiction of sexually explicit conduct, (3) illegal drug or alcohol use, (4) self-harm or suicide, or (5) a crime of violence against another person. Violations carry BOTH criminal exposure (RSA 639:3 Endangering the Welfare of a Child, para III-a) and civil liability: a private track (RSA 507:8-k para. III — actual damages, with at least $1,000-per-violation liquidated damages, brought by a child, parent, or next friend; damages only, no injunctive relief and no attorney's-fees provision) and a separate AG track (RSA 507:8-k para. II — damages or appropriate injunctive relief, only after written notice + 90-day cure). Carve-outs exist for cloud/telecom/information-service providers (third-party content) and for AI chat programs or characters integral or incidental to a video game, TV, streaming, or movie experience.
Deadline: January 1, 2026
RSA 639:3(III-a); RSA 507:8-k(II)-(III)Age-Appropriate AI Content Filtering
High PriorityImplement content moderation and response filtering that identifies and blocks age-inappropriate AI outputs. Maintain audit logs demonstrating the chatbot does not generate prohibited harmful content for users who may be minors. Review and update filters as AI capabilities evolve.
Deadline: January 1, 2026
RSA 639:3(III-a); RSA 507:8-kWho Does This Apply To?
Applies to: any person that provides a "responsive generative communication" — responsive open-ended generative communication through AI (an AI chat program, large language model bot, chat bot, character AI, or other computer application) — accessible to a child (under 18) and available to New Hampshire residents (eff. Jan 1, 2026). The duty is to not knowingly/intentionally use it to communicate with a child with intent to facilitate, encourage, offer, solicit, or recommend that the child imminently engage in: (1) sexually explicit conduct; (2) the production of, or participation in, a visual depiction of sexually explicit conduct; (3) illegal use of drugs or alcohol; (4) acts of self-harm or suicide; or (5) a crime of violence against another person. No business-size threshold — the trigger is operating a child-accessible generative-communication service, not the size of the operator. CARVE-OUTS: (a) a cloud-service provider, or a telecommunications-service or information-service provider, with respect to content created by or supplied on behalf of another person; and (b) any product/service/website/application that provides an AI chat program or character integral or incidental to a video game, television, streaming, movie, or other similar interactive experience. TWO enforcement tracks: (1) CRIMINAL — the conduct amends RSA 639:3 (Endangering the Welfare of a Child), new para III-a; and (2) CIVIL, itself split into two separate sub-tracks within new RSA 507:8-k — (2a) a PRIVATE track (para. III) giving a child, parent, or next friend the right to recover actual damages proximately caused by a violation, with a liquidated-damages floor of at least $1,000 per violation, but NO injunctive relief and no attorney's-fees/fee-shifting provision; and (2b) an ATTORNEY GENERAL track (para. II, textually separate from the private track) under which the AG may seek damages (per para. III) or appropriate injunctive relief, but only after written notice and a 90-day opportunity to cure. CYCLE 23 (2026-08-23): corrected — injunctive relief and the notice/cure requirement attach only to the AG track, not the private track; confirmed against the verbatim RSA 507:8-k paragraph I-IV text (via gencourt.org and fastdemocracy.com, cross-checked, byte-identical quotes) after gc.nh.gov and law.justia.com again refused/403'd this cycle exactly as they did in Cycle 22.
Recent Regulatory Guidance
New Hampshire HB 143 (Chapter 270, 2025) — AI child-safety duties; criminal + civil; AG enforcement
HB 143 (Chapter 270; signed 1 Aug 2025 by Gov. Ayotte; effective 1 Jan 2026) — its AI provision (substance of SB 263) bars using a "responsive generative communication" to communicate with a child with intent to facilitate, encourage, offer, solicit, or recommend that the child imminently engage in sexually explicit conduct (incl. producing/participating in a visual depiction of it), illegal drug/alcohol use, self-harm or suicide, or a crime of violence. It runs on two tracks: a CRIME (amending RSA 639:3, Endangering the Welfare of a Child, para III-a) AND a civil track split across new RSA 507:8-k para. III (private right of action for a child, parent, or next friend: actual damages + at least $1,000-per-violation liquidated damages, NO injunctive relief, no attorney's-fees provision) and para. II (AG-only track: damages or appropriate injunctive relief, only after written notice and a 90-day cure). No separate AG compliance bulletin has been published — the statutory text sets the requirements.
Key Case Law & Precedent
Garcia v. Character Technologies, Inc. (M.D. Fla. 2024)
US District Court, Middle District of Florida · 2024Wrongful-death case alleging Character.AI's chatbot contributed to a 14-year-old's suicide. NH legislative findings reference Garcia as the leading case demonstrating the harm pattern HB 143 targets. The bill's child-safety prohibitions (no facilitation of self-harm, no encouragement of harmful acts) are direct legislative responses to the conduct alleged in Garcia.
Outcome: Settled in principle 2026-01-07 (with 4 related family suits in CO/NY/TX); financial terms not disclosed; Character.AI/Google committed to additional under-18 safety features. (Cycle 8, 2026-08-22: corrected from a stale "litigation active" claim already identified and fixed in nebraska_lb525 by Cycle 6 but left unfixed here — same fabrication surviving in a sibling jurisdiction entry.)
Case referenceFrequently Asked Questions
Does New Hampshire HB 143 — AI Chatbot Child Safety Law apply to my business?
New Hampshire HB 143 (Chapter 270, signed August 1, 2025 by Gov. Ayotte, effective January 1, 2026) — an omnibus act whose AI provision (substance of SB 263) prohibits a person from using a "responsive generative communication" (an AI chat program,… Use Aegis Firma's free scanner to get a personalized assessment in under 5 minutes.
What is the penalty for non-compliance?
The maximum penalty under New Hampshire HB 143 — AI Chatbot Child Safety Law is: Criminal: a violation of RSA 639:3 (Endangering the Welfare of a Child), new para III-a (no separate class designated within III-a; prosecuted within the endangering-welfare framework). Civil, private track (RSA 507:8-k para. III): actual damages proximately caused by a violation, with a liquidated-damages floor of at least $1,000 per violation, brought by a child, parent, or next friend — damages only, no injunctive relief and no attorney's-fees provision in this paragraph. Civil, AG track (RSA 507:8-k para. II, separate from the private track): the Attorney General may seek damages (per para. III) OR appropriate injunctive relief, but only after written notice and a 90-day opportunity to cure.. Fines are typically scaled by company size, severity of violation, and whether violations were willful or accidental.
How do I comply with New Hampshire HB 143 — AI Chatbot Child Safety Law?
The 2 requirements above cover the core obligations. The fastest path to compliance is: (1) conduct an AI risk assessment, (2) document your AI systems, (3) implement transparency disclosures where required. Aegis Firma generates all required documents automatically.
Official Source
https://www.citizenscount.org/bills/hb-143-2025Last updated: 2026-08-23 — verify at source before relying on this information.
Don't leave compliance to chance
Aegis Firma scans your AI tools, tells you exactly which regulations apply, and generates all required documents — in 30 minutes.
Start your free compliance scan