S. 5117 — Senior Chatbot Protection Act of 2026

This section defines key terms for the Act. It establishes an artificial intelligence chatbot as an interactive service or application that generates non-predetermined, adaptive responses to open-ended natural-language or multimodal input, excluding narrow-purpose tools limited to predetermined replies. It also defines affirmative consent as requiring clear, standalone disclosures with equal ease of refusal; crisis as expressions of suicidal ideation, intent to harm, or medical emergencies; high-stakes decisions as those affecting health, safety, finances, or independence with potentially irreversible consequences; older adult as an individual age 65 or older; and related terms including covered entity, service provider, and crisis service provider.

This section requires artificial intelligence chatbots to disclose at conversation start and regular intervals that they are not human, prohibits false claims of being licensed professionals, and mandates detection of high-stakes decisions with disclosures that outputs are not professional advice and lack legal privilege. It further requires crisis detection with referrals to the 988 Suicide and Crisis Lifeline, bars advice on self-harm or medication, limits collection and processing of conversation data except for service delivery or legal compliance, mandates affirmative consent for model training, and requires user deletion of conversation history.

This section requires covered entities to submit annual reports to the National Institutes of Health Director and the Commission on material adverse incidents involving artificial intelligence chatbots and high-stakes decisions or crises, including totals disaggregated by incident type, user age groups (under 50, 50–64, 65–74, 75–84, and 85 and older), and crisis response outcomes such as referrals and follow-up actions. It directs the Commission to issue reporting standards within 180 days and requires annual public release of aggregate, de-identified data.

This section authorizes the Federal Trade Commission to enforce sections 3 and 4 as unfair or deceptive acts or practices under the Federal Trade Commission Act, using the same powers, jurisdiction, and remedies. It requires the Commission to give covered entities written notice and 60 days to cure a first violation before seeking civil penalties, except for intentional deception, substantial imminent harm, or repeated violations; imposes civil penalties of up to $50,000 per knowing or reckless violation; empowers state attorneys general to bring parens patriae actions with notice to the Commission; and preserves enforcement of stronger state laws.

This section directs the Director of the National Institute of Standards and Technology to develop voluntary guidelines within 18 months for artificial intelligence chatbots interacting with older adults. The guidelines address transparency and disclosure, accessibility and age-appropriate design, identification of crisis situations with referrals to resources, and cybersecurity and privacy safeguards. Covered entities must consider the guidelines when designing and deploying such chatbots. The Director must consult with the Federal Trade Commission, National Institute on Aging, National Institute of Mental Health, industry representatives, consumer advocates, older adults, caregivers, and specified experts.

This section directs the Commission to promulgate rules within one year after enactment describing the form, content, timing, and frequency of disclosures required under section 3(a)(1), providing example templates for those disclosures, specifying the form and manner for reports required under section 4, and establishing other requirements to carry out the Act except section 6.

This section provides that if any provision of the Act is held invalid, the remainder of the Act and its application to other persons or circumstances remains in effect.

This section preserves all existing rights, claims, remedies, presumptions, and defenses under consumer protection, privacy, and civil rights laws by providing that the Act does not preempt or otherwise affect them.