Consumer Protection - Consumer Reporting Agencies - Use of Algorithmic Systems

HB 1477 requires consumer reporting agencies (like credit bureaus) that use algorithmic systems to generate credit reports to meet specific transparency and accuracy standards. It mandates agencies to explain algorithmic decisions in plain language, maintain a public registry of all algorithms used, and achieve strict thresholds: less than 0.5% error rate in algorithmic evaluations compared to human review, under 0.1% discriminatory bias based on protected characteristics, and 99.9% data accuracy. Agencies must also conduct quarterly independent bias audits, implement data governance frameworks with minimum data requirements (1,000-10,000 data points per category), and ensure human review of automated decisions within 24 hours. This bill directly affects entities that use algorithms to assess consumer credit for third-party reporting in Maryland, aiming to reduce bias and increase accountability in automated credit scoring.