Deploys a companion chatbot, in the United States

15 U.S.C. Section 45 · Read the law's page

• Do not engage in unfair or deceptive practices when collecting, using, or sharing personal data.
• Assess foreseeable privacy harms before deploying a system that collects or uses biometric identifiers derived from photographs, videos, or voice recordings, and disclose material facts about that collection.

15 U.S.C. 7001(a), (c) and (h) · Read the law's page

• Treat a contract or record that your electronic agent forms or delivers in a transaction in or affecting interstate or foreign commerce as one that cannot be denied effect solely because an electronic agent acted; it binds the person to whom the agent's action is legally attributable (15 U.S.C. 7001(h), 7006(3)).
• Before you rely on an electronic record to give a consumer information that a law requires to be in writing, obtain the consumer's affirmative consent after a clear and conspicuous statement of the consumer's rights to paper and to withdraw consent, whether the consent covers one transaction or categories of records, the withdrawal procedure, and the hardware and software needed, and take the consent or its confirmation electronically in a manner that reasonably demonstrates the consumer can access the electronic form (15 U.S.C. 7001(c)(1)(A) to (C)).
• If a change in the hardware or software needed to access or retain electronic records creates a material risk that the consumer cannot access or retain a later record, give the consumer a statement of the revised requirements and of the right to withdraw consent without fees, and obtain consent again (15 U.S.C. 7001(c)(1)(D)).
• Where a law requires you to retain a contract or record, keep an electronic record that accurately reflects the information in it and stays accessible, in a form capable of being accurately reproduced, to everyone entitled to access it for the required period (15 U.S.C. 7001(d)(1)).

16 CFR Part 461 · Read the law's page

• Do not materially and falsely pose, directly or by implication, as a government entity or an officer, executive, official, employee or agent of one, in or affecting commerce.
• Do not materially and falsely pose, directly or by implication, as a business, including a not-for-profit entity, or as an officer, executive, official, employee or agent of one, in or affecting commerce.
• Do not materially misrepresent, directly or by implication, your affiliation with a government entity or a business or an officer of either, including endorsement or sponsorship by it, in or affecting commerce.

No. 6:24-cv-1903-ACC-UAM (M.D. Fla. May 21, 2025)

• Treat the design of a conversational AI app as open to product liability claims: on a motion to dismiss, the court held the app a product so far as claims arise from defects in the app, such as the alleged failure to confirm users' ages, omission of reporting mechanisms, programming of characters to employ human mannerisms, and users' inability to exclude indecent content, rather than from ideas or expressions within it.
• Do not assume a chatbot's output is protected speech: at the pleading stage the court was not prepared to hold that the output of the large language model behind the app is speech.
• Take precautions against foreseeable harm to users: at the pleading stage the court held that releasing the chatbot to the public created a foreseeable risk of harm that the developer was in a position to control, which sufficed to allege a duty.
• Do not present chatbot characters as real people or as licensed mental health professionals: at the pleading stage the court held that allegations that characters insist they are real people and are labeled as psychologists or therapists pleaded deceptive conduct under Florida's deceptive and unfair trade practices statute.

Discussion Draft, 119th Cong. (June 4, 2026)

H.R.1694, 119th Cong. (2025-2026)

H.R.6489, 119th Cong. (2025-2026)

• This measure is in committee and binds nobody yet; what follows is what it would require if it is enacted in this form.
• Disclose at the start of each interaction that the chatbot is an AI system, not a human professional.
• Provide that disclosure in age-appropriate plain language when the user is a minor under 17.
• Provide crisis resource referrals.
• Send a break reminder after three hours of use.
• Adopt a policy against harmful content.

H.R.7985, 119th Cong. (2025-2026)

• This measure is in committee and binds nobody yet; what follows is what it would require if it is enacted in this form.
• Do not imply that a chatbot's output is provided or verified by a licensed human professional, if you operate in a licensed profession such as finance, insurance, or health care.

S. 4199, 119th Congress · Read the law's page

• Introduced and ordered reported by committee, not enacted, so this binds nobody yet; what follows is what it would require as introduced.
• If you make an AI chatbot available and know a user is under 18, disclose at the beginning of every session, and at least every 30 minutes, that the user is not interacting with a human and that the chatbot's content is AI-generated.
• Do not configure an AI chatbot, for a user you know is under 18, with rewards or incentives based on frequency of use, push notifications, outputs the user did not prompt, or typing indicators, once the Federal Trade Commission's regulations prohibit them.
• Do not configure an AI chatbot to advertise to a user you know is under 18, and do not process that user's personal data to profile the user or to train a covered algorithm, other than to test for and address risks of harm.
• Do not retaliate against an individual who raises a concern about, or reports, a violation of the Act.

S. 4407, 119th Congress · Read the law's page

• Introduced and ordered reported by committee, not enacted, so this binds nobody yet; what follows is what it would require as introduced.
• If you run a public-facing website, online service or app whose primary function is an AI chatbot, require a user you know is under 13 to create and maintain a family account to use it.
• Before a teen you know is 13 to 17 creates an account, give a parent direct notice and obtain verifiable parental consent.
• Give each family account parental controls to limit time spent, and to disable rewards, notifications, financial transactions and unprompted outputs, and let the parent access a full record of the child's or teen's conversations and activity.
• Do not use the personal data of a user you know is a child or teen for targeted advertising.
• Terminate the existing account of a known child without a family account, or of a known teen whose parent has not given verifiable consent, and delete the personal data collected from that user.

S.1213 / H.R.5272, 119th Cong. (2025-2026)

• This measure is in committee and binds nobody yet; what follows is what it would require if it is enacted in this form.
• Do not distribute materially deceptive AI-generated audio, image, or video depicting a federal candidate in a political ad.

S.2164, 119th Cong. (2025-2026)

• This measure is in committee and binds nobody yet; what follows is what it would require if it is enacted in this form.
• Conduct an impact assessment of an automated decision system used in a high-stakes context such as housing, employment, credit, or education.
• Disclose the impact assessment.
• Tell consumers when an automated decision system is being used to make a decision about them.
• Let consumers opt out of the use of an automated decision system.

S.4591, 119th Cong. (2025-2026)

• This measure is in committee and binds nobody yet; what follows is what it would require if it is enacted in this form.
• Do not create or use an unauthorized AI-generated digital replica of a person's voice or visual likeness.
• Do not host an unauthorized AI-generated digital replica of a person's voice or visual likeness.