Congress mandates AI chatbot warnings for older adults, crisis detection

What it does

This bill requires AI chatbot companies to clearly disclose when users are talking to a bot (not a human), warn users before they rely on chatbots for major life decisions (medical, legal, financial, estate planning), detect and refer users in crisis to the 988 Suicide Lifeline, and delete user conversation history on request. It imposes special protections for older adults, restricts companies from using manipulative design to trap users in conversations, and requires annual reporting of serious incidents to the FTC and NIH. The FTC enforces violations as unfair trade practices, with civil penalties up to $50,000 per violation.

Why we flagged it

The bill's operative mechanism is a set of mandatory disclosures, safety protocols, and data-protection rules for AI chatbot operators, enforced by the FTC. It is fundamentally a consumer-protection statute, not a subsidy, deregulation, or industry carve-out.

What the text implies

• The bill's definition of 'high-stakes decision' is broad and includes estate planning, guardianship, and financial decisions—areas where older adults are vulnerable to manipulation. The FTC's rulemaking authority to expand this category (Section 2(7)) could significantly widen compliance obligations if new categories are added.
• Crisis detection relies on 'evidence-based methods' and NIST guidance (Section 3(1)(D)), but the bill does not specify what those methods are or how accurate they must be. A chatbot that fails to detect a crisis, or falsely detects one, may face liability, creating incentive to over-refer and potentially desensitize users to crisis warnings.
• The 'affirmative consent' standard for training on user data (Section 3(2)(3)) is strict: users must opt-in before their conversations are used for model training. This may limit companies' ability to improve chatbots using real-world interaction data, potentially slowing innovation in safety and accuracy.
• Reporting requirements (Section 4) are disaggregated by age group (under 50, 50–64, 65–74, 75–84, 85+), creating a detailed public record of which age groups experience material adverse incidents. This transparency is a public benefit but may also expose companies to targeted litigation or regulatory scrutiny if older-adult incident rates are high.
• The bill does not create a private right of action—only the FTC and state attorneys general can enforce it. Users harmed by a chatbot cannot sue directly under this statute, limiting individual remedies to state consumer-protection laws or common law.

Section numbers refer to the bill text the analysis read — linked under Primary records below.

Who it affects

Ordinary users gain concrete protections: mandatory disclosure that they're talking to a bot, warnings before relying on chatbots for high-stakes decisions, crisis detection and referral, and the right to delete their conversation history. Older adults receive heightened safeguards against manipulative design. The bill does not restrict access to chatbots or impose broad liability shields on companies—it sets transparency and safety floors. The 60-day cure period before penalties is reasonable a

Named in the bill

Federal Trade Commission (FTC), National Institute of Standards and Technology (NIST), National Institutes of Health (NIH), National Institute on Aging, National Institute of Mental Health, 988 Suicide and Crisis Lifeline, State attorneys general, AI chatbot operators (covered entities)

Where it stands

1 cosponsor: 1 Republicans.

• Jul 23, 2026 — Introduced · Congress.gov: “Introduced in Senate”
• Jul 23, 2026 — Referred to Senate Committee on Commerce, Science, and Transportation · Congress.gov: “Read twice and referred to the Committee on Commerce, Science, and Transportation”

Dates and quoted wording are Congress.gov's action record; the timeline shows status changes, not every procedural step.