Skip to content
AI Policy Generator

ESWhat applies

Companion or social chatbot in United States (federal): related rules

Chatbots designed to sustain personal relationships or conversations, such as friendship, emotional support or characters. Several US states already require notices, crisis protocols and protections for minors.

Risk classification

JurisdictionLevelDateNote
CaliforniaTransparencyJanuary 1, 2026SB 243: AI notice, suicidal ideation protocol, reminders and safeguards for minors; annual report from Jul 1, 2027; private right of action.
New York (incl. NYC)TransparencyNovember 5, 2025New York AI companion law (General Business Law Article 47), in force since Nov 5, 2025: crisis protocol and periodic notice that it is not a person.
WashingtonTransparencyJanuary 1, 2027HB 2225: mandatory safeguards with heightened protection for minors; excludes customer service bots; private right of action.
NebraskaTransparencyJuly 1, 2027LB 525 (Conversational AI Safety Act): operational duties from July 2027 (exact date to be verified).
UtahSector rulesMay 7, 2025HB 452: mental health chatbots must disclose they are AI; filing a written policy provides a legal defense.
United States (federal)Data protectionApril 22, 2026Amended COPPA Rule if the service is directed to children under 13; the FTC is examining companion chatbots used by minors.

Obligations related to this use case

Disclose that the chatbot is AI

California SB 243Cal. Bus. & Prof. Code § 22602(a), (d)In force

Applies from January 1, 2026

When it relates
Applies to companion chatbots with users in California where a reasonable person could believe they are talking to a human. Excludes bots used only for customer service or business operations, video game characters limited to the game, and voice assistants that do not sustain a relationship.
What the rule provides
  • Show a clear notice that the user is talking to AI if a reasonable person could think they are talking to a human.
  • State in the app or website that companion chatbots may not be suitable for some minors.
Typical evidence
  • Screenshots of the in-product notice
Verification
Pending confirmation in the official source (last checked: November 4, 2026)leginfo.legislature.ca.gov returned 403 on Nov 4, 2026. The section numbers (§§ 22601-22605), the three-hour reminder for minors and the annual report from Jul 1, 2027 are still unchecked in the official text. Signing and Chapter 677 come from consistent secondary sources (Skadden, Jones Walker, LegiScan).

Protocol for suicidal ideation and self-harm

California SB 243Cal. Bus. & Prof. Code § 22602(b)In force

Applies from January 1, 2026

When it relates
Applies to companion chatbots with users in California.
What the rule provides
  • Maintain a protocol that prevents suicide or self-harm content and refers users to crisis lines when they express it.
  • Publish the protocol details on the website.
Typical evidence
  • Published protocol
  • Log of crisis-service referrals
Verification
Pending confirmation in the official source (last checked: November 4, 2026)leginfo.legislature.ca.gov returned 403 on Nov 4, 2026. The section numbers (§§ 22601-22605), the three-hour reminder for minors and the annual report from Jul 1, 2027 are still unchecked in the official text. Signing and Chapter 677 come from consistent secondary sources (Skadden, Jones Walker, LegiScan).

Safeguards for minor users

California SB 243Cal. Bus. & Prof. Code § 22602(c)In force

Applies from January 1, 2026

When it relates
Applies when the operator knows the user is a minor.
What the rule provides
  • Remind the user that it is AI and suggest a break at least every three hours of continued use.
  • Take reasonable measures so the chatbot does not produce sexually explicit content or encourage minors to engage in it.
Typical evidence
  • Reminder settings and content filters
Verification
Pending confirmation in the official source (last checked: November 4, 2026)leginfo.legislature.ca.gov returned 403 on Nov 4, 2026. The section numbers (§§ 22601-22605), the three-hour reminder for minors and the annual report from Jul 1, 2027 are still unchecked in the official text. Signing and Chapter 677 come from consistent secondary sources (Skadden, Jones Walker, LegiScan).

Annual report to the Office of Suicide Prevention

California SB 243Cal. Bus. & Prof. Code § 22603Adopted, applies later

Applies from July 1, 2027

When it relates
Applies to companion chatbot operators with users in California.
What the rule provides
  • Report annually on the number of crisis-service referrals and the protocols in place, without personal data.
Typical evidence
  • Annual report submitted
Verification
Pending confirmation in the official source (last checked: November 4, 2026)leginfo.legislature.ca.gov returned 403 on Nov 4, 2026. The section numbers (§§ 22601-22605), the three-hour reminder for minors and the annual report from Jul 1, 2027 are still unchecked in the official text. Signing and Chapter 677 come from consistent secondary sources (Skadden, Jones Walker, LegiScan).

Protocol for suicidal ideation or self-harm

New York AI companionsN.Y. Gen. Bus. Law § 1701In force

Applies from November 5, 2025

When it relates
Applies to AI companions with users in New York State.
What the rule provides
  • Do not offer the AI companion without a protocol that detects expressions of suicidal ideation or self-harm and refers the user to crisis services.
Typical evidence
  • Approved protocol and detection and referral tests
Verification
Verified in the official source on November 4, 2026Checked on nysenate.gov: § 1701 (protocol with reasonable efforts to detect and address suicidal ideation or self-harm and a notice referring users to crisis services such as 9-8-8) and § 1703 (Attorney General; up to $15,000 per day). The effective date (Nov 5, 2025) was checked on Nov 2, 2026 in the article index and the Governor's letter.

Notify that it is not a person at the start and periodically

New York AI companionsN.Y. Gen. Bus. Law § 1702In force

Applies from November 5, 2025

When it relates
Applies to AI companions with users in New York State.
What the rule provides
  • Clearly notify the user, verbally or in writing, that they are not communicating with a human at the start of the interaction (this notice need not be repeated more than once per day) and at least every three hours if the interaction continues.
Typical evidence
  • Notice settings and screenshots
Verification
Verified in the official source on November 4, 2026Checked on nysenate.gov: § 1702 requires a clear notice, verbal or written, at the start of the interaction (no more than once per day is needed) and at least every three hours if the interaction continues. Penalties in § 1703.

Confirm it is AI when the consumer asks

Utah AI Policy ActUtah Code § 13-75-103 (según reforma SB 226)In force

Already applicable

When it relates
Applies if you use generative AI to interact with consumers in Utah.
What the rule provides
  • Clearly answer that it is AI and not a human when the consumer clearly asks.
  • To use the safe harbor, disclose clearly at the start of and throughout the interaction.
Typical evidence
  • Chatbot disclosure script or settings
Verification
Pending confirmation in the official source (last checked: November 4, 2026)le.utah.gov refused the connection on Nov 4, 2026, so the official text is still unchecked. The wording after SB 226 still has to be confirmed in Utah Code § 13-75-103, and § 63I-2-213 has to show whether the Jul 1, 2027 repeal also covers what SB 226 added. Content per FPF and Perkins Coie.

AI disclosure and data and advertising limits for mental health chatbots

Utah HB 452Utah Code § 13-72a-201 y ss. (HB 452, 2025)In force

Already applicable

When it relates
Applies to generative AI chatbots that a reasonable person would believe can provide mental health therapy, with users in Utah. The assessment detects it through chatbot plus healthcare; it does not apply to clinical tools that do not converse or to scripted content (for example, guided meditations).
What the rule provides
  • Disclose that it is AI and not a human before access, when the user returns after a period of inactivity and when the user asks.
  • Do not sell or share identifiable health data or user inputs.
  • Label advertising and do not use what the user says to target ads.
Typical evidence
  • In-product notices
  • Data and advertising policy
Verification
Pending confirmation in the official source (last checked: November 4, 2026)le.utah.gov refused the connection on Nov 4, 2026, so the official text is still unchecked. The final section numbering, the disclosures and the written-policy defense still have to be confirmed in Utah Code Title 13, Chapter 72a. Content per Davis Wright Tremaine, FPF, Sidley and WSGR.

Disclose AI at the start and periodically

Washington HB 2225ESHB 2225, sec. 2 (cap. 168, Leyes de 2026)Adopted, applies later

Applies from January 1, 2027

When it relates
Applies to companion chatbots with users in Washington.
What the rule provides
  • Clearly disclose that the chatbot is artificial and not human at the start and at least every three hours.
  • Prevent the chatbot from claiming to be human or contradicting the notice.
Typical evidence
  • Notice settings and functional tests
Verification
Verified in the official source on November 2, 2026Signing, chapter and effective date checked on app.leg.wa.gov; content per the official final bill report.

Self-harm protocol and public reporting

Washington HB 2225ESHB 2225, sec. 3Adopted, applies later

Applies from January 1, 2027

When it relates
Applies to companion chatbots with users in Washington.
What the rule provides
  • Do not offer the chatbot without a protocol that detects suicidal ideation or self-harm and refers users to crisis resources.
  • Prevent content that encourages or describes self-harm.
  • Publish on the website and in the app the protocol and the prior year's number of referrals.
Typical evidence
  • Published protocol and annual referral count
Verification
Verified in the official source on November 2, 2026Content per the official final bill report (lawfilesext.leg.wa.gov).

Enhanced protections for minors

Washington HB 2225ESHB 2225, sec. 2Adopted, applies later

Applies from January 1, 2027

When it relates
Applies if the operator knows the user is a minor or the chatbot is directed to minors.
What the rule provides
  • Repeat the AI notice every hour of continued use.
  • Prevent sexually explicit content or suggestive dialogue with minors.
  • Prohibit manipulative techniques designed to create or prolong an emotional relationship.
Typical evidence
  • Filter and notice settings for minors
Verification
Verified in the official source on November 2, 2026Content per the official final bill report.

Disclose that the user is talking to AI

Nebraska LB 525LB 525, Conversational Artificial Intelligence Safety ActAdopted, applies later

Applies from July 1, 2027

When it relates
Applies to publicly accessible conversational AI services with account holders in Nebraska. The definition may reach customer service chatbots: confirm the exclusions in the official text.
What the rule provides
  • Clearly and conspicuously disclose that the user is interacting with AI when a reasonable person could believe they are talking to a human.
  • Do not present the service as professional mental or behavioral health care.
Typical evidence
  • In-product notices
Verification
Pending confirmation in the official source (last checked: November 4, 2026)nebraskalegislature.gov refused the connection on Nov 4, 2026, so the signed law (LB 525 slip law) is still unchecked: operative date (Jul 1, 2027), definitions, notices to minors, self-harm protocol, penalty amounts and whether customer service bots are excluded. Signing (Apr 14, 2026) and content per consistent secondary trackers.

Additional safeguards for minors

Nebraska LB 525LB 525, Conversational Artificial Intelligence Safety ActAdopted, applies later

Applies from July 1, 2027

When it relates
Applies to operators with account holders under 18.
What the rule provides
  • Show minors a persistent notice that it is AI, or a notice at the start of each session and at least every three hours of continuous interaction.
  • Do not give minors points or similar rewards at unpredictable intervals to increase engagement.
Typical evidence
  • Minors protocol
Supporting documents
Verification
Pending confirmation in the official source (last checked: November 4, 2026)nebraskalegislature.gov refused the connection on Nov 4, 2026, so the signed law (LB 525 slip law) is still unchecked: operative date (Jul 1, 2027), definitions, notices to minors, self-harm protocol, penalty amounts and whether customer service bots are excluded. Signing (Apr 14, 2026) and content per consistent secondary trackers.

Protocol for suicidal ideation or self-harm

Nebraska LB 525LB 525, Conversational Artificial Intelligence Safety Act (secs. 12-18)Adopted, applies later

Applies from July 1, 2027

When it relates
Applies to companion chatbots with users in Nebraska; it may reach other conversational services depending on the official text.
What the rule provides
  • Adopt a protocol to respond to prompts about suicidal ideation or self-harm, with reasonable efforts to refer users to crisis services.
Typical evidence
  • Approved protocol and detection and referral tests
Verification
Pending confirmation in the official source (last checked: November 4, 2026)nebraskalegislature.gov refused the connection on Nov 4, 2026, so the signed law (LB 525 slip law) is still unchecked: operative date (Jul 1, 2027), definitions, notices to minors, self-harm protocol, penalty amounts and whether customer service bots are excluded. Signing (Apr 14, 2026) and content per consistent secondary trackers.

Related base documents

Related kits

Includes 5 documents

Customer Service Chatbot Kit

AI interaction notice, right to reach a human, explanations, complaints and incidents for a customer-facing chatbot or voice assistant.

USD 109USD 175

Complete kit · available to buy

Includes 5 documents

Education and Minors Kit

Chatbot protocol for minors and crises, DPIA, AI notices, human oversight in assessment and a check for prohibited practices such as emotion recognition.

USD 109USD 175

Complete kit · available to buy

Find out what applies to you (free)United States (federal)