This session examines executive agencies’ use of AI in administrative adjudications. Executive agencies adjudicate thousands of claims every year, attempting to balance efficiency and fairness. Often, agencies’ adjudication systems are overwhelmed and bogged down in delays, errant decisions, and backlogs. Millions of claimants, including veterans, immigrants, social security beneficiaries and others wait years for decisions that are wrongly decided. Now, agencies are looking to AI to alleviate backlogs and streamline adjudication processes.
This panel will examine the role of AI in administrative decision-making and how it affects the diagnosis and litigation of issues in clinical practice. The panelists are experienced veterans law attorneys and have litigated claims before the Department of Veterans’ Affairs (VA), though the issues they will discuss are broadly applicable to any area of law that relies on (or will soon rely on) the use of AI in adjudications. Using VA as a case study, the panelists will discuss the potential for AI mishaps in mass adjudications. Part I of the presentation will address authorities governing the use of AI, as well as basic principles upon which AI algorithms work. Panelists will also examine the various ways that administrative agencies have already started employing AI in their adjudications. Part II will examine the apparent benefits and potential pitfalls of automation in agency decision-making. Part III will focus on teaching students to understand automation bias, to identify AI in their respective practice areas, and to consider the impact of AI on client counseling and client representation generally. Attendees will have the opportunity to share their experiences with AI and to discuss how students might learn to identify and respond to this new agency tool.