EPIC Testifies in Support of Connecticut Bill on Algorithmic Discrimination

EPIC Law Fellow Kara Williams testified today in support of Connecticut S.B. 2, which seeks to regulate the development and use of high-risk AI systems.

EPIC’s testimony focuses on why bills like S.B. 2 are urgently needed because they address harms from companies using high-risk AI systems in making important decisions about people’s lives, including whether they get a job or admission into a university, if they qualify for a mortgage or a loan, how much they pay for rent or insurance, and whether they can access health care or government services.

Specifically, EPIC supports the bill’s transparency and individual rights provisions. These measures will provide individuals with more information about AI systems and how they are used by companies that are making decisions about them.

S.B. 2’s goal of addressing algorithmic discrimination is an important one, but EPIC suggests some amendments to further strengthen the bill to ensure it meets this goal. EPIC’s recommended amendments include: (1) changing the duty of care to avoid algorithmic discrimination to an outright prohibition on algorithmic discrimination; (2) adding a private right of action into the bill’s enforcement structure; (3) clarifying the definitions of “substantial factor” and “high-risk artificial intelligence system” to ensure that they do not contain loopholes that would allow companies to self-select out of complying with the bill; and (4) removing the exception to individuals’ right to appeal adverse decisions.

EPIC has worked with state legislators across the country on similar AI bills and is eager to continue this advocacy.