Sponsors (5)
The bill creates clearer, time‑bound pathways for supervised AI experimentation in finance and stronger data protections for sponsors, but it also raises the risk of enforcement gaps and new compliance burdens—while leaving insurance consumers outside these agency protections.
Financial institutions and consumers get clearer legal definitions and regulator assignment for AI in financial products, improving accountability and reducing uncertainty about which agency oversees AI-related financial decisions.
Financial institutions can run supervised AI test projects under agency oversight, enabling faster, safer innovation and potential efficiency gains in financial products and services.
Applicants (mostly financial institutions) receive a time‑bound review process (120 days plus one extension), so they get faster decisions and reduced regulatory uncertainty.
Consumers, taxpayers, and financial institutions face heightened risk that regulatory gaps, overlapping jurisdiction, or automatic approvals (if agencies miss deadlines) will allow risky AI projects to proceed or result in uneven enforcement.
Smaller regulated firms and some financial institutions bear new compliance, monitoring, and reporting costs to classify, test, and supervise AI projects, which may disadvantage small businesses.
Consumers who rely on insurance products may be less protected because the business of insurance is excluded from the bill's definition of covered 'financial product or service.'
Based on analysis of 3 sections of legislative text.
Requires federal financial regulators to create AI Innovation Labs allowing approved AI test projects under alternative compliance strategies, with set timelines, data protections, and reporting.
Creates AI Innovation Labs inside each major federal financial regulator and lets regulated firms run approved AI test projects under an alternative compliance plan. Agencies must set application rules, decide applications quickly, provide limited non‑enforcement protections for approved tests, secure supplied data, and file annual anonymized reports on outcomes for several years.
Introduced July 29, 2025 by Marion Michael Rounds · Last progress July 29, 2025