ARTICLE
30 July 2025

Massachusetts AG Settles With Student Loan Lender On AI-Based Fair Lending Violations

SM
Sheppard, Mullin, Richter & Hampton LLP

Contributor

Businesses turn to Sheppard to deliver sophisticated counsel to help clients move ahead. With more than 1,200 lawyers located in 16 offices worldwide, our client-centered approach is grounded in nearly a century of building enduring relationships on trust and collaboration. Our broad and diversified practices serve global clients—from startups to Fortune 500 companies—at every stage of the business cycle, including high-stakes litigation, complex transactions, sophisticated financings and regulatory issues. With leading edge technologies and innovation behind our team, we pride ourselves on being a strategic partner to our clients.
On July 10, Massachusetts Attorney General Andrea Joy Campbell announced a $2.5 million settlement with a student loan company to resolve allegations that its underwriting practices...
United States Massachusetts Technology
Sheppard, Mullin, Richter & Hampton LLP are most popular:
  • within Cannabis & Hemp and Insolvency/Bankruptcy/Re-Structuring topic(s)

Listen to this post

On July 10, Massachusetts Attorney General Andrea Joy Campbell announced a $2.5 million settlement with a student loan company to resolve allegations that its underwriting practices violated the Massachusetts Consumer Protection Act and the Equal Credit Opportunity Act, including through the use of artificial intelligence models that produced disparate impacts on protected groups.

The Attorney General's investigation focused on both automated and manual underwriting systems used to issue and refinance student loans. Regulators alleged that the company's models and override processes lacked appropriate oversight, failed to assess for discriminatory effects, and incorporated inputs that may have disproportionately harmed non-white and non-citizen applicants. The matter was resolved through an assurance of discontinuance filed in Suffolk County Superior Court.

Specifically, the settlement alleged that the company:

  • Failed to test its AI models for fair lending risks. The company allegedly used algorithmic underwriting tools without evaluating them for disparate impact or applying adequate controls.
  • Used school-level default data that hurt minority applicants. The models reportedly factored in cohort default rates tied to specific colleges, which resulted in less favorable outcomes for certain racial groups.
  • Provided denial notices that lacked accurate reasons. The lender allegedly issued adverse action notices that failed to explain the real basis for credit denials due to system design limitations.
  • Allowed underwriters to override models without controls. Manual overrides allegedly occurred without written policies, creating inconsistencies in outcomes for similarly situated applicants.
  • Lacked oversight over its underwriting practices. The company allegedly failed to implement policies, testing, or documentation to ensure compliance with state and federal law.

The settlement requires the company to implement AI governance procedures, conduct annual fair lending testing, eliminate certain underwriting variables, and report compliance efforts to the Attorney General's Office.

Putting It Into Practice: States continue to expand their enforcement actions in response to perceived gaps in federal oversight (previously discussed here and here). Massachusetts is among the first states to impose both monetary penalties and extensive injunctive relief tied to alleged disparate impact and compliance failures. This is especially notable in light of the federal government's discontinuance of disparate impact analysis in bringing far housing and fair lending cases (previously discussed here).

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.

[View Source]
See More Popular Content From

Mondaq uses cookies on this website. By using our website you agree to our use of cookies as set out in our Privacy Policy.

Learn More