WeVote
Log In

Bill

Bill

SB 2195

AN ACT RELATING TO COMMERCIAL LAW--GENERAL REGULATORY PROVISIONS -- ARTIFICIAL INTELLIGENCE COMPANION MODELS

2026 Regular Session Introduced by Pete Appollonio and 6 co-sponsors

Rhode Island would regulate AI companions to require safety protocols, clear non-human disclosures, and enforcement to prevent self-harm, harm to others, or financial harm.

06/22/2026 Signed by Governor
0
WeVote Research Nonpartisan
Bill Summary · SB 2195

Overview

SB 2195, introduced in Rhode Island for the 2026 session, would regulate AI companion models under the state’s commercial law framework. The bill establishes definitions, requires safety and transparency measures, sets notification obligations, and provides enforcement mechanisms. It takes effect January 1, 2027.

Main purpose and intent

  • To create safety standards and consumer protections for AI companions—AI systems designed to simulate social interaction and provide advice related to personal well-being.
  • To ensure users are aware they are interacting with a non-human computer program and to address potential harms (self-harm, harm to others, financial harm) communicated through AI companions.
  • To empower the state’s AG and consumer protection authorities to enforce these rules and to maintain ongoing coordination via a state AI task force.

Key provisions and changes

1) Definitions (6-63-1)

  • Introduces “AI companion,” defined as an AI system that simulates social interaction, remembers prior interactions, asks questions, and engages in conversations about personal well-being. Excludes systems primarily providing information about products/services or customer accounts.
  • Defines related terms: “AI model,” “Artificial intelligence,” “Emotional recognition algorithms,” “Generative AI,” “Operator,” and “User.”

2) Prohibitions and requirements (6-63-2)

  • Operators may not operate or provide an AI companion unless the system includes a protocol to address:
    • Possible suicidal ideation or self-harm expressed by a user.
    • Possible physical harm to others expressed by a user.
    • Possible financial harm to others expressed by a user, including directing the user to crisis services if relevant.

3) Notifications required (6-63-3)

  • Operators must give a notification at the start of an AI interaction and at least every three hours for ongoing interactions.
  • Notification content (displayed verbally or in bold, 16-point type) must state: “THE AI COMPANION (OR NAME OF THE AI COMPANION) IS A COMPUTER PROGRAM AND NOT A HUMAN BEING. IT IS UNABLE TO FEEL HUMAN EMOTION.”

4) Enforcement (6-63-4)

  • Civil remedies: Individuals harmed by violations (self-harm, physical, or financial harm) may sue in Superior Court for damages and equitable relief.
  • AG authority: The attorney general can investigate, sue, and seek injunctions against noncompliant AI providers, including action under deceptive trade practices laws.

5) Consultation (6-63-5)

  • The Rhode Island AI Taskforce will continue advising the governor and General Assembly on AI development, implementation, and regulation.

6) Severability (6-63-6)

  • If any provision is found unconstitutional or invalid, the rest of the chapter remains in effect.

Who/what is affected

  • Operators (providers) of AI companion models offering services within Rhode Island.
  • Users of AI companions in the state.
  • The Rhode Island Attorney General’s office (enforcement and investigations).
  • The Rhode Island AI Taskforce (ongoing advisory role).

Timeline and procedural aspects

  • Effective date: January 1, 2027.
  • Action history indicates committee consideration in 2026, with scheduled readings and potential amendments prior to enactment.

Potential impact and considerations

  • Increased transparency: Clear disclosures that AI companions are non-human could affect user trust and adoption.
  • Safety measures: Protocols for self-harm, physical harm, and financial harm could require operators to implement safety workflows and crisis referrals.
  • Compliance burden: Operators may need changes to interfaces to include mandatory notifications and threat/harm assessment protocols.
  • Legal exposure: Provides pathways for civil suits and AG enforcement, potentially deterring unsafe AI practices.
  • Industry effect: Could influence the design of AI companion products sold or used in Rhode Island, possibly affecting available features and user experience.

Compiled from official sources — confirm details with the bill’s official record.

Sign in to ask a question.