The "Unleashing AI Innovation in Financial Services Act" mandates that each financial regulatory agency establish an AI Innovation Lab . These labs are designed to enable regulated entities to experiment with artificial intelligence (AI) test projects in financial products or services, with reduced regulatory burden and without the expectation of enforcement actions for specific regulations. Regulated entities can submit applications detailing their AI test projects, which must make substantial use of AI and fall under agency jurisdiction. The application requires an alternative compliance strategy , identifying regulations to be waived or modified, and proposing how the entity will effectively manage risks. It must also explain how the project serves the public interest, improves consumer access, fosters innovation, enhances efficiency, or improves risk management, while ensuring no systemic risk, compliance with anti-money laundering obligations, or national security risks. Agencies have 120 days to review applications, with a possible 120-day extension, after which an application is automatically deemed approved if no determination is made. Upon approval, the agency agrees to enforce the specified regulation only as outlined in the alternative compliance strategy, and other financial regulatory agencies generally cannot enforce that particular regulation. However, agencies retain authority to take enforcement actions for fraud or unsafe practices related to an AI test project. If an application is denied, the agency must provide a written explanation, and the entity may resubmit amended applications, though no more than two substantially similar ones. Agencies can seek injunctive relief if an AI test project poses an immediate danger to consumers or investors, or presents a significant risk to financial markets, federal insurance funds, AML/CFT obligations, or national security. The bill also allows for joint applications from multiple entities or projects spanning multiple agencies, requiring joint approval. Within 180 days of enactment, each financial regulatory agency must promulgate regulations outlining procedures for modifying projects, consequences for non-compliance, project termination (not earlier than one year), extensions, confidentiality, and coordination for joint applications. Furthermore, each agency must submit annual reports for seven years, detailing aggregated findings, trends, and lessons learned from AI test projects, without disclosing proprietary information or the names of participating entities.
Unleashing AI Innovation in Financial Services Act
USA119th CongressS-2528| Senate
| Updated: 7/29/2025
The "Unleashing AI Innovation in Financial Services Act" mandates that each financial regulatory agency establish an AI Innovation Lab . These labs are designed to enable regulated entities to experiment with artificial intelligence (AI) test projects in financial products or services, with reduced regulatory burden and without the expectation of enforcement actions for specific regulations. Regulated entities can submit applications detailing their AI test projects, which must make substantial use of AI and fall under agency jurisdiction. The application requires an alternative compliance strategy , identifying regulations to be waived or modified, and proposing how the entity will effectively manage risks. It must also explain how the project serves the public interest, improves consumer access, fosters innovation, enhances efficiency, or improves risk management, while ensuring no systemic risk, compliance with anti-money laundering obligations, or national security risks. Agencies have 120 days to review applications, with a possible 120-day extension, after which an application is automatically deemed approved if no determination is made. Upon approval, the agency agrees to enforce the specified regulation only as outlined in the alternative compliance strategy, and other financial regulatory agencies generally cannot enforce that particular regulation. However, agencies retain authority to take enforcement actions for fraud or unsafe practices related to an AI test project. If an application is denied, the agency must provide a written explanation, and the entity may resubmit amended applications, though no more than two substantially similar ones. Agencies can seek injunctive relief if an AI test project poses an immediate danger to consumers or investors, or presents a significant risk to financial markets, federal insurance funds, AML/CFT obligations, or national security. The bill also allows for joint applications from multiple entities or projects spanning multiple agencies, requiring joint approval. Within 180 days of enactment, each financial regulatory agency must promulgate regulations outlining procedures for modifying projects, consequences for non-compliance, project termination (not earlier than one year), extensions, confidentiality, and coordination for joint applications. Furthermore, each agency must submit annual reports for seven years, detailing aggregated findings, trends, and lessons learned from AI test projects, without disclosing proprietary information or the names of participating entities.