On August 14, CDT submitted comments in response to the House Financial Services Committee’s request for information on the application of current federal laws and regulations related to the use of AI in the financial marketplace and any proposed reforms to provide for an appropriately modernized and comprehensive federal financial AI framework.
The Committee and regulators have explored implications of AI adoption in recent years, and CDT has called for stronger enforcement of laws and regulations and recommended reforms to promote accountability of AI integrated into the provision of financial services, housing, and related needs. Our comments:
- Suggest requirements to improve model fairness, explainability, and model risk management;
- Identify areas for improvement in existing financial data privacy laws and the limitations of disclosures alone as a governance measure;
- Discuss liability of financial institutions and their third-party service providers, and financial institutions’ obligations to manage their third-party services; and
- Recommend steps to address AI’s impacts on housing and workers.
Facts Only
Executive Summary
Full Take
The interaction highlights a structural tension between technological adoption in finance and the lagging development of regulatory and governance structures. The recommendations focus on shifting the locus of responsibility—demanding requirements for fairness, explainability, and risk management within the AI model itself, alongside clarifying liability across complex institutional relationships (institutions versus third-party providers). This points toward a pattern where technological advancement outpaces legal consensus, creating governance vacuums in areas like data privacy and social impact. The suggestions to address housing and worker impacts suggest an acknowledgment that financial AI is not purely a technical or fiduciary issue but carries broader socio-economic consequences. The exercise reveals a systemic challenge: how to build accountability frameworks when the systems are complex, multi-layered, and evolving faster than traditional legal structures can adapt. The missing link in many regulatory discussions is often the integration of external social costs and explicit governance mandates into core compliance mechanisms.
Bridge Questions: What specific metrics or standards would be necessary to operationalize "model fairness" across diverse financial products? How can liability frameworks be designed to effectively distribute responsibility between primary institutions and embedded service providers? What proactive, preventative regulatory mechanisms are needed to address externalities impacting housing stability and labor markets before harm materializes?