Public Health - Ibogaine Clinical Research Grant Program - Establishment (Veterans Mental Health Innovations Act)
HB1477 would add a new section to Maryland’s consumer reporting law regulating consumer reporting agencies that use algorithmic systems to assemble or evaluate consumer credit information. The bill applies to agencies that use automated tools to generate consumer reports and requires them to be able to explain algorithmic evaluations in plain language, maintain a public registry of algorithms and their data sources, and meet specified performance standards for error rates, data accuracy, and discriminatory outcomes.
The bill also requires quarterly independent harmful-bias audits, annual reporting to the Commissioner of Financial Regulation, and a data governance framework that tracks data lineage, certifies sources, and uses minimum dataset sizes for statistical reliability. It further requires human review of automated evaluations within 24 hours, an expedited review process, random human review of at least 10% of evaluations, alternative nonalgorithmic assessment options for consumers who opt out, and contingency planning for system failures or breaches. The Commissioner would be directed to set annual assessment thresholds, require regular training for human reviewers, and create a whistleblower protection program, with authority to adopt implementing regulations.
HB1477 would expand Maryland’s consumer protection and credit reporting rules by imposing detailed compliance obligations on consumer reporting agencies that rely on algorithmic systems. It would affect agencies that furnish consumer reports for credit, insurance, employment, or other authorized purposes, and would create new oversight responsibilities for the Commissioner of Financial Regulation under the Commercial Law Article. The bill would likely require affected businesses to invest in auditing, documentation, human review procedures, bias testing, and consumer-facing explanation systems, while giving regulators new tools to monitor algorithmic decision-making in credit reporting.
Based on the bill text and the available legislative context, the measure appears to be framed as a consumer-protection and transparency bill aimed at reducing errors and harmful bias in automated credit reporting. There are no recorded committee transcripts or votes in the provided materials, so there is no direct evidence of support or opposition from hearings or floor action. The bill’s structure suggests a strong regulatory approach, which may appeal to consumer advocates and fairness-focused policymakers while drawing concern from industry stakeholders about feasibility and compliance costs.
The main points of contention are likely to be the bill’s strict technical standards and operational requirements. Potentially disputed provisions include the very low allowable error and discriminatory-data thresholds, the requirement for public disclosure of algorithms and methodologies, mandatory human review within short timeframes, and the 10% random human review mandate. Consumer reporting agencies and data/credit industry groups may argue these requirements are costly, difficult to implement, or could expose proprietary systems, while supporters would likely emphasize transparency, accountability, and protection against automated bias and inaccurate credit decisions.