What does Preamble of AI LEAD Act do?

This bill aims to establish legal standards for advanced artificial intelligence products.

What does FINDINGS. of AI LEAD Act do?

This section outlines findings related to artificial intelligence (AI) systems and their impact. The text states that AI systems shift decision-making power from humans to software. It acknowledges that AI products have caused and will likely cause harm to businesses and individuals, citing the tragic deaths of teenagers exploited by AI chatbots. The findings express concern that unpredictable liability allocations threaten public safety and the financial stability of individuals and industries. Product liability law is identified as a potential tool to address harms caused by AI systems affecting interstate commerce. The Act aims to establish a federal products liability framework for AI systems to promote innovation, protect due process rights, and remove barriers to interstate commerce. The legislation intends to provide predictable legal outcomes and promote business innovation while avoiding implications for expressive speech.

What does DEFINITIONS. of AI LEAD Act do?

This section defines key terms used throughout the Act. An 'artificial intelligence system' is defined as software or a data system capable of making predictions or decisions using machine learning or other algorithmic methods. A 'claimant' is any person bringing a liability action. A 'covered product' is defined as an artificial intelligence system. A 'deployer' is a person who uses or operates a covered product, either for personal/commercial use or for a third party. The 'design' of a covered product includes its material characteristics, formulation, data used for training, and the training/testing processes. A 'developer' is a person who designs, codes, produces, owns, or substantially modifies a covered product. Harm is broadly defined to include damage to property, physical injury, financial/reputational injury, and mental/psychological anguish. A 'liability action' is a civil action brought under section 301 based on harm caused by a covered product. A 'substantial modification' is a deliberate change made to a covered product that alters its purpose or function, and is not authorized by the developer.

What does DEVELOPER LIABILITY FOR HARM TO BUSINESS OR CONSUMER. of AI LEAD Act do?

This section establishes liability standards for developers of covered products. A claimant must prove the developer's failure to exercise reasonable care in design, instructions/warnings, warranty fulfillment, or product condition proximately caused the harm. For defective design claims, a claimant must prove a reasonable alternative design existed that could have reduced harm, unless the design is manifestly unreasonable. Circumstantial evidence may be used to infer a product defect if the harm is of a kind typically resulting from such defects. Noncompliance with applicable safety statutes or regulations can deem a product defective. Inadequate instructions or warnings render a product defective if reasonable instructions/warnings could have reduced harm. Developers are not liable for failure to warn about open and obvious risks, except for users under 18. Developers can be strictly liable for harm caused by a defective product, even without contractual relationship with the claimant or exercising all possible care. Developers are not liable for harm solely caused by a substantial modification to the product.

What does DEPLOYER LIABILITY FOR HARM TO BUSINESS OR CONSUMER. of AI LEAD Act do?

This section establishes deployer liability for harm caused by a covered product, linking it to developer liability under section 101. A deployer can be deemed liable if they make a substantial modification to the covered product or intentionally misuse it contrary to its intended use. A use intended by the developer is not considered a substantial modification or misuse. If a developer doesn't specify an intended use, it will be inferred by the targeted market and manner of distribution. Deployers licensing a covered product are not liable for violations of section 101 solely due to ownership or use, subject to section 302.

What does UNCONSCIONABLE LIABILITY LIMITATIONS. of AI LEAD Act do?

This section prohibits developers and deployers from including contract language that waives rights, restricts forums or procedures, or unreasonably limits liability related to harm caused by covered products. Such language in contracts with deployers is deemed unenforceable. Similarly, developers and deployers are prohibited from including such language in terms and conditions relevant to covered products. Language in terms and conditions violating these prohibitions is also unenforceable.

What does FEDERAL CAUSE OF ACTION. of AI LEAD Act do?

This section establishes a federal cause of action. The Attorney General, state attorneys general, individuals, or their legal representatives, or a class of individuals may bring a civil action. The civil action can be brought against a developer or deployer for violations of sections 101, 102, or 201. Available relief includes injunctive relief, civil penalties (if brought by the Attorney General), damages, restitution, or other compensation for individuals. Plaintiffs may also recover reasonable attorney fees and litigation costs. The Attorney General or a State attorney general may seek other appropriate relief.

What does SPECIAL RULE FOR DEPLOYERS. of AI LEAD Act do?

This section establishes a special rule for 'deployers' in liability actions related to harm. If a developer is not subject to a court's jurisdiction, is insolvent, or unable to answer for harm, a deployer may be held liable to the same extent as the developer. A court must dismiss a deployer from a liability action if the developer is a party, subject to the court's jurisdiction, not insolvent, and the deployer is not otherwise liable under section 102. If both the developer and deployer contributed to the harm, they may be jointly and severally liable for the portion of harm caused by their respective conduct. A deployer held liable for harm caused by a developer may seek indemnification, including attorney fees and litigation costs, from the developer. Indemnification for a deployer is limited if they are at fault for a portion of the harm. This section preserves the claimant's right to maintain a liability action against the developer, the deployer, or both.

What does PERIOD OF LIMITATIONS. of AI LEAD Act do?

This section establishes a four-year statute of limitations for liability actions. The statute of limitations begins when a claimant discovers, or should have discovered with reasonable care, both the harm and its cause. The statute of limitations is tolled for individuals under a legal disability until they cease to be so disabled. The statute of limitations is also tolled from the date of a complaint filing against a developer or deployer until a final judgment is entered in the case.

What does PREEMPTION. of AI LEAD Act do?

This Act establishes a preemption clause regarding state law. State law is superseded by this Act only where a conflict exists. States may enact or enforce stronger protections for covered products than those provided by this Act, provided they align with harm prevention, accountability, and transparency.

Back to bills

AI LEAD Act

Last updated: 9/29/2025 · Introduced: 9/29/2025

Author: Richard Durbin (D-IL)

TL;DR (AI)

This bill establishes a federal framework for product liability related to artificial intelligence (AI) systems, aiming to balance innovation with consumer and business protection.
It defines key terms like 'artificial intelligence system,' 'developer,' and 'deployer,' clarifying responsibilities for AI-related harm.
Developers can be held liable for harm caused by defective AI systems if they fail to exercise reasonable care in design, instructions, or warranties.
Deployers (users or operators of AI systems) can be liable if they substantially modify the system or misuse it contrary to the developer’s intended use.
The bill prohibits contracts that limit liability related to AI harm and establishes a four-year statute of limitations for related lawsuits, while allowing states to enact stronger protections.

119th Congressin committeeOfficial page

Verified Votes

No verified votes yet.

Community Votes

No community votes yet. Be the first!

Voting as guest · Sign in for verified votes