EPIC Releases its AI Legislation Scorecard

Today, EPIC published its AI Legislation Scorecard, a first-of-its-kind rubric for lawmakers, journalists, advocates and academics to use when evaluating the strength of state and federal AI bills. EPIC’s AI Legislation Scorecard comes as the United States faces a tidal wave of new AI legislation, with hundreds of AI bills being introduced in at least 40 states and dozens of federal regulations following suit.

To ground its review of the growing AI legislative landscape, EPIC set out to create a structured tool for evaluating AI bills. Informed by expert consultations, internal bill analysis, and evidence-backed policy research, EPIC’s AI Legislation Scorecard outlines key provisions that effective AI legislation should contain, including but not limited to data minimization requirements, impact assessment and testing obligations, prohibitions on particularly harmful AI uses, and robust enforcement mechanisms.

The launch of EPIC’s AI Legislation Scorecard was paired with an expert panel discussion involving Vermont State Representative Monique Priestley; Nik Marda, Mozilla’s Technical Lead for AI Governance; Alicia Solow-Niederman, Associate Professor at the George Washington University Law School; and Adam Billen, Director of Policy at Encode Justice. As Nik Marda explained, EPIC’s Scorecard “shows that we actually know a lot about how to regulate AI already. While [responsible AI oversight] is framed like a hopeless endeavor, we’ve actually made a lot of progress through the years on how to tackle [AI] harms.” With this Scorecard, EPIC provides a guide for what robust, comprehensive AI legislation can look like.

If you have questions or would like EPIC to grade a specific piece of state or federal AI legislation, please contact EPIC Law Fellow Kara Williams.