On February 12, the House Financial Services Committee’s Task Force on Artificial Intelligence held a hearing titled “Equitable Algorithms: Examining Ways to Reduce AI Bias in Financial Services.” Members engaged the panel in a technical conversation about methods being employed by machine learning experts to mitigate the impacts of bias in the use and deployment of AI tools. Members on both sides of the aisle acknowledged that there were important tradeoffs between the accuracy of AI models and the fairness of the outcomes they inform. Various tactics to address bias in AI were explored, including auditing and validating models, maintaining human oversight in the development and use of models, maintaining diverse legal and engineering teams and innovating algorithm designs to better isolate biased model components. BPI submitted a letter for the record, which highlighted the promise AI holds for improving fairness for decisions regarding consumer credit.
You Might Also Be Interested In...
COVID-19 Relief BPI and Joint Coalition Urge House and Senate Leadership to Address PPP Forgiveness Process
COVID-19 Relief Loan Necessity Questionnaires -Borrower & Lender Coalition Letter to Congressional Leadership
Consumer Affairs Loan Necessity Questionnaires – Borrower & Lender Coalition Letter to SBA & Treasury
FinTech & Innovation FinTech and Big Tech Companies Want the Benefits of Banking Without the Responsibilities. Regulatory Loopholes Could Let Them Succeed.
FinTech & Innovation Is It OK for FinTechs To Use Regulatory Arbitrage To Avoid Consolidated Supervision?
More Posts by This Author
Bank Capital and Stress Testing GSIB Method 2 Fixed Coefficients Must Be Adjusted for Economic Growth
AML, Bank Secrecy Act and Sanctions BPI Endorses Legislative Compromise to Reform Anti-Money Laundering Framework and End Anonymous Shell Companies
Bank Activities and Structure BPI Asks OCC to Furnish Data Used to Support Rushed Fair Access Proposal in FOIA Request