Idaho SB 1297, the Conversational AI Safety Act, is a state law regulating public-facing conversational AI services, codified at Title 48, Idaho Code (Consumer Protection), as a new Chapter 21 — short title at §48-2101, definitions at §48-2102. It imposes disclosure requirements, a suicide-referral protocol, and a prohibition on impersonating mental-health professionals, with additional protections for minors covering disclosure cadence, engagement design, sexual content, and anthropomorphism. The Act takes effect July 1, 2027.
The bill was passed by the Sixty-eighth Idaho Legislature, Second Regular Session — 2026, and was sponsored by the State Affairs Committee.
Scope and definitions
The Act applies to conversational AI services — AI applications, web interfaces, or programs accessible to the general public that primarily simulate human conversation (textual, visual, or aural). It excludes developer and researcher tools, narrow-discrete-topic systems, commercial-business-use products, voice-assistant speakers, internal-business-only systems, customer-service chatbots, and video-game chatbots that do not engage on mental health, self-harm, or topics harmful to minors.
A minor is defined as anyone under 18 where the operator has actual knowledge or reasonable certainty of that status.
Operator requirements (§48-2103)
Operators of covered services are subject to three obligations. The first is disclosure: where reasonable persons would be misled to believe they are interacting with a human, the operator must clearly and conspicuously disclose the AI nature of the service. The second is a suicide protocol requiring operators to refer users expressing suicidal ideation to crisis services. The third prohibits mental-health-professional impersonation: an operator may not knowingly cause the service to make any representation explicitly indicating it is designed to provide professional mental or behavioral health care.
Minor-specific requirements (§48-2104)
Where the operator knows or is reasonably certain a user is a minor, additional requirements apply:
- Disclosure cadence: a persistent visible disclaimer, or disclosure at the beginning of each session and every three (3) hours.
- No variable rewards: the service may not use variable rewards at unpredictable intervals to encourage engagement.
- Sexual-content prohibitions: the operator must prevent visual material of sexually explicit conduct, statements urging engagement in such conduct, or sexual objectification of the minor. "Sexually explicit conduct" and "visual depiction" take the meanings given in 18 U.S.C. §2256.
- Anti-anthropomorphism prohibitions: the operator must prevent statements that simulate emotional dependence, romantic or sexual innuendo, role-play of adult-minor romantic relationships, or explicit claims of sentience or human nature.
- Privacy and account tools: these are required for minors and, for users under 13, their parents or guardians.
Enforcement and penalties (§48-2105)
The Act sets a civil penalty of $1,000 per violation, capped at $500,000 per operator, or actual damages, whichever is greater. Enforcement is reserved to the Attorney General; the Act contains no private right of action.
A developer carve-out provides that the chapter does not create liability for the developer of an AI model for any violation by a conversational AI system made available to the public by a third-party operator.
Comparative position
| Provision | Idaho SB 1297 | [[legislation/nebraska-lb525 | NE LB525]] | [[legislation/oregon-sb-1546 | OR SB 1546]] | [[legislation/washington-hb-2225 | WA HB 2225]] | [[california-sb-243\ | CA SB 243]] | [[legislation/tennessee-sb-1580 | TN SB 1580]] |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Effective | 2027-07-01 | 2027-07-01 | 2027-01-01 | 2027-01-01 | 2026-01-01 | 2026-07-01 | |||||
| Scope | Capability | Capability | Behavior (narrowest) | Capability | Capability | "AI system" (broad) | |||||
| Patient-care carve-out | No | No | Yes | No | No | n/a | |||||
| Disclosure cadence (minors) | 3 hr OR persistent | 3 hr OR persistent | 3 hr | 1 hr | Per CA framework | n/a | |||||
| 8 manipulative engagement techniques | No | No | 3 techniques | Yes (8 enumerated) | Limited | n/a | |||||
| Sexual-content prohibition (minors) | Yes | Yes | Yes | Yes | Yes | n/a | |||||
| Mental-health-prof impersonation | Yes | Yes | No | No | No | Yes (sole focus) | |||||
| PRA | No | No | Yes ($1K stat. dmg.) | Yes (CPA) | Yes | No | |||||
| Per-violation cap | $1K / $500K | $1K / $500K | $1K stat. dmg. | CPA penalties | Per CA framework | $5K | |||||
| Developer carve-out | Yes | Yes | n/a | n/a | n/a | n/a |
Relationships
- related: California SB 243 — Companion Chatbots, Nebraska LB525 — Conversational AI Safety Act + Agricultural Data Privacy Act, Oregon SB 1546 — Artificial Intelligence Companions, Washington ESHB 2225 — AI Companion Chatbots, Tennessee SB 1580 / HB 1470 — AI Mental-Health Professional Impersonation Prohibition — 2026 wave of state companion-chatbot and mental-health AI laws.
- depends-on: AI Mental Health and Psychological Harm — concept-level case the statute instantiates.
- related: Character.AI Litigation (umbrella) — the Idaho framework addresses the harms alleged in those cases (impersonation of professionals, romantic adult-minor role-play, sexually explicit content).
- supports: Techno-Federalism — Idaho is part of the Mountain West cluster joining the West-Coast and Mountain regulatory front.
Source
Primary text: Raw Sources/Idaho Conversational AI Safety Act (SB 1297).md (source_class: foundational). Secondary analysis: The Rest of the West: Oregon and Washington Build on California Chatbot Law (covers Oregon and Washington; the Idaho framework slots into the same multi-state grid).