This bill creates "AI Innovation Labs" inside each federal financial regulatory agency. Regulated financial firms can apply to test financial products or services that use artificial intelligence. Applications must describe the AI project, propose an alternative way to meet specific agency regulations, explain how the project serves the public interest or protects consumers or investors, limit size and growth, give a business plan, and estimate economic impact. Agencies must review applications within 120 days (with one possible 120-day extension). If an agency approves an application, the agency is generally limited to enforcing the relevant regulation only as described in the approved alternative compliance strategy for the length of the test project; other agencies generally may not enforce that regulation for the project. Agencies must adopt rules for the program within 180 days after enactment, including confidentiality, modification procedures, and minimum test length rules. Agencies must keep submitted data secure. Agencies must report aggregated findings and lessons learned starting 2 years after enactment and each year for 7 years after that. The bill keeps agency authority to act against fraud, unsafe or unsound practices, threats to financial markets, deposit insurance loss, anti-money laundering violations, or national security risks, and it allows agencies to seek court injunctions when projects present those dangers.
If you are a bank, broker, investment adviser, credit union, or other entity supervised by a federal financial regulator and you want to test an AI-powered financial product or service, this bill would let you apply to run a limited test under an agency-approved alternative compliance plan. Approved tests would run under terms set by the approving agency and include limits on size and duration. Agencies will publish rules for how tests operate and will report aggregated results to Congress. The bill does not remove agency power to act against fraud or serious risks.
No publicly available information.
The bill requires applicants to explain how their AI test project would serve the public interest, improve consumer or investor access or protection, enhance efficiency, foster innovation or competitiveness, improve risk management or security, or enhance regulatory compliance. The bill allows regulated entities to try AI projects under defined conditions and alternative compliance strategies, which supporters can point to as a way to test innovation while giving agencies oversight and reporting on outcomes.
No publicly available information.