US Federal Senate Bill SB2081

US Federal 2025-2026 Regular Session

• Introduced: 6/12/25
• Refer: 6/12/25

Caption

RISE Act of 2025

Summary

The RISE Act of 2025 would create a conditional safe harbor from civil liability for certain artificial intelligence developers when their AI products are used by licensed or otherwise authorized professionals in the course of providing services to clients. The bill is aimed at professional settings such as healthcare, law, finance, and other regulated fields, and it is premised on the idea that clearer liability rules and greater transparency will encourage responsible AI adoption while preserving accountability for serious misconduct. To qualify for immunity, a developer would have to publicly release and continuously maintain a model card and model specification before deployment, and provide clear documentation to professional users about known limitations, failure modes, and appropriate uses. The bill also requires updates to those materials within 30 days after a new version is deployed or a new material failure mode is discovered. The immunity would not apply to reckless or willful misconduct, and it would not protect developers from state-law claims based on fraud, knowing misrepresentation, or conduct outside the scope of professional use. The bill would take effect on December 1, 2025, and would apply only to acts or omissions on or after that date.

Impact

The bill would add a federal liability framework for AI developers and expressly preempt state-law claims against covered developers for errors arising from AI use by learned professionals, but only when the bill’s transparency and documentation conditions are met. It would not eliminate all liability; claims for fraud, knowing misrepresentation, and non-professional use would remain available, and other federal, state, or common-law immunities would be preserved. The measure would therefore affect developers, professional users, and clients in sectors where AI is used to support expert judgment, while also influencing how AI products are documented, marketed, and updated.

Sentiment

There is no recorded committee debate or vote history in the provided materials, so no direct legislative sentiment can be measured from hearings or roll calls. The bill text itself reflects a pro-innovation but accountability-focused approach, emphasizing responsible integration of AI, transparency, and risk management. Overall, the measure appears designed to reassure developers and professional users by clarifying liability exposure, while also signaling concern about AI errors and the need for public disclosure of model behavior.

Contention

The main policy tension in the bill is between shielding AI developers from broad civil liability and preserving remedies for harmed clients and other plaintiffs. Supporters would likely favor the safe harbor as a way to reduce uncertainty and encourage adoption of AI in professional services, while critics may view the immunity as too protective of developers, especially if disclosures are incomplete or redacted for trade-secret reasons. Another likely point of contention is the scope of the transparency requirements, including what must be disclosed in model cards and model specifications, how trade secrets are protected, and whether the 30-day update window is sufficient after new failure modes are discovered.