Capstone Practicum by Keertan Kini, Kevin Li, Sigalit Perelson, and Eyal Zilberman

Our student practicum team was invited by the California Privacy Protection Agency (CPPA) to analyze legislative and regulatory acts governing automated decisionmaking (ADM) and profiling at the federal level, in California, and in other U.S. states. Definitions of automated decisionmaking vary, and automated decisionmaking has been the subject of various regulatory efforts in the United States and internationally. In this paper, we examine approaches by U.S. federal agencies; California legislation spanning employment, housing, and child protection; and privacy and data protection statutes in other U.S. states.

We conclude that the various examined legislative and regulatory acts attempt to reduce ambiguity by enumerating harms that may require intervention, but they diverge in their level of specificity and applicability. No single definition of automated decisionmaking has been broadly adopted. Substantive regulatory requirements differ due to varying policy goals pursued and constraints faced by different legislative and regulatory authorities.