Rhode Island House Bill 8052 is a pending bill of the January Session 2026 that would add a new section 9-1-55 to chapter 9-1 of the General Laws, "Causes of Action." It imposes liability on developers of artificial intelligence systems, regardless of the degree of care exercised, for injuries factually and proximately caused by a system engaging in conduct that would be negligent, an intentional tort, or a crime if undertaken by an adult human of sound mind. The bill was introduced on February 27, 2026 by Representatives Carson, Caldwell, Knight, and Cotter, and referred to House Judiciary (Rhode Island H 8052 — Artificial Intelligence: Liability for Injuries (bill text)).
The Legislative Council's explanation of the bill reads in full: "This act would provide a civil cause of action for individuals injured by artificial intelligence."
Status and timeline
| Date | Event |
|---|---|
| 2026-02-27 | Introduced; referred to House Judiciary |
The bill has not advanced out of committee. It would take effect upon passage. Sponsors are identified in the bill header by surname only; full names are not established from the document.
Scope
The bill reaches any artificial intelligence system — "an engineered or machine-based system that varies in its level of autonomy and that can, for explicit or implicit objectives, infer from the input it receives how to generate outputs that can influence physical or virtual environments." No compute threshold, cost threshold, or capability threshold appears anywhere in the text, so the regime is not confined to frontier models.
A developer is a person who performs the initial training of a system "either by training a model, or by fine-tuning an existing model." Fine-tuning therefore makes a party a developer and brings it within the liability rule.
Key provisions
Liability rule (§ 9-1-55(b)). Except for defamation claims, developers "shall be liable, regardless of the degree of care they exercised, for all injuries" that are factually and proximately caused by an AI system engaging in conduct that would be negligent, an intentional tort, or a crime if an adult human of sound mind engaged in it, and where that conduct "was neither intended nor a consequence of negligence on the part of" the user or an intermediary that fine-tuned, scaffolded, or otherwise modified the system. There is no user/non-user distinction, so a user of the system may recover.
Mental-state presumption (§ 9-1-55(c)). For torts where mental state is an element, a rebuttable presumption arises that the system held the relevant mental state where the finder of fact concludes by a preponderance that a natural person acting similarly in similar circumstances would have. The rebuttal mechanics are stated separately for jury and bench trials, with the burden of persuasion resting on the party against whom the presumption is invoked in both. Subsection (c)(3) provides that "it shall not be a defense that artificial intelligence systems are incapable of having mental states."
Absolute defenses (§ 9-1-55(d)). Two defenses are provided, each styled "an absolute defense to liability, including liability under the common law or existing statutes": that the developer establishes the model satisfied the standard of care applicable to humans performing the same function, or that the injuries resulted from a "capabilities failure" — the model falling short of intended or reasonably anticipated performance without engaging in conduct that would be tortious or criminal if a person did it. Because the defenses expressly displace common-law and existing statutory liability, a developer who establishes either is insulated beyond the new section (Rhode Island H 8052 — Artificial Intelligence: Liability for Injuries (bill text)).
An undefined term in the operative text
The liability rule at subsection (b) attaches to an "artificial intelligence system," which subsection (a) defines. Both defenses at subsection (d) attach instead to a "covered model or covered model derivative," neither of which subsection (a) defines; the phrase appears only inside the definition of fine-tuning. Read literally, the defenses may not reach the conduct the liability rule covers. The New York near-twin bill defines "covered model" and uses the term consistently, but importing that definition would be an inference rather than a reading of this text (Rhode Island H 8052 — Artificial Intelligence: Liability for Injuries (bill text)).
Comparison with New York A8833
New York A8833, introduced June 9, 2025, shares this bill's AI definition, its rebuttable-presumption mechanism, its bar on the no-mental-states defense, and both defenses, in closely parallel language. The comparison table is at that page. The three divergences that most change the regime are that H 8052 has no compute or cost threshold where A8833 reaches only models above 10^26 operations and $100 million in training cost; that H 8052 permits users as well as non-users to recover; and that H 8052's defenses are absolute and displace common-law and existing statutory liability where A8833's are affirmative defenses to its own rule. Whether the two descend from a common model text is not established by either document.
The combination in H 8052 — no threshold, users may recover, defenses reach existing law — makes it simultaneously the broader liability rule and the broader immunity, depending on which side establishes its case.
Relationships
- depends-on: Rhode Island H 8052 — Artificial Intelligence: Liability for Injuries (bill text) — the bill text.
- related: New York A8833 — Understanding Artificial Intelligence Act — near-twin bill in New York.
- instance-of: AI Liability — a statutory liability regime for AI system conduct.
- supports: AI and Tort Liability — a legislative alternative to developing the theory through common-law litigation.
- related: State-Level AI Regulation.