Bankers Urge Principles-Based Approach to AI Regulation; Preemption of State Law

The American Bankers Association ("ABA") urged the White House Office of Science and Technology Policy ("OSTP") to adopt a risk-based, innovation-supportive approach to regulating artificial intelligence ("AI"). 

In a comment letter responding to OSTP’s Request for Information on AI regulatory reform, the ABA emphasized that federally regulated financial institutions already operate under robust supervisory and risk management standards established by the Administration’s AI Action Plan. The Association noted that these standards address a full range of AI-related risks, including model governance, fair lending, cybersecurity, and third-party oversight. The ABA stated that new rules should enhance flexibility, promote responsible innovation, and avoid creating overlapping or inconsistent requirements for the banking industry.

The ABA recommended:

1. Update Supervisory Guidance on Model Risk Management. The ABA urged the Federal Reserve, OCC, and FDIC to modernize their joint model risk management guidance to (i) reflect current AI technologies, including generative AI, (ii) clarify responsibilities for third-party AI models, (iii) establish a taxonomy for AI tools to scope the guidance appropriately, and (iv) focus on outcomes rather than prescriptive methodologies for validation.
2. Reform Bank Field Examinations. The ABA called for examination reforms to ensure supervisors focus on substantive risk factors—such as model inputs, outputs, and outcomes—rather than technical minutiae. The Association noted that overly prescriptive or inconsistent examiner practices can inhibit innovation and push financial activities toward less-regulated entities.
3. Encourage Voluntary Standards and AI Centers of Excellence. The ABA supported the creation of voluntary, industry-led standards and certifications, such as standardized model disclosure templates ("model cards"), to enhance transparency and accountability. The Association proposed establishing AI Centers of Excellence to coordinate best practices, develop shared risk management tools, and align industry efforts with the NIST AI Risk Management Framework.
4. Enact Comprehensive Federal AI Legislation. The ABA urged Congress to pass federal AI legislation establishing a unified risk management framework with strong preemption of conflicting state laws. The Association emphasized that any federal standards should acknowledge the banking sector’s existing regulatory structure to prevent duplicative oversight and ensure continued leadership in responsible AI adoption.