ABA Webinar on Developing Effective AI Compliance Programs Summarized by Analysis Group Consultants
May 14, 2024
Analysis Group Vice President David Toniatti and Analyst Georgina Scoville have published a recap of the American Bar Association (ABA) webinar “Designing Effective AI Compliance Programs” on the ABA’s Antitrust Law Section website. The two economists summarize a discussion among moderator Rishi Satia (Morgan Lewis) and panelists Amir Ghavi (Fried Frank), Ben Rossen (OpenAI), and John Horton (MIT Sloan School of Management). The webinar was cohosted by the ABA’s Joint Conduct Committee and Compliance and Ethics Committee.
The panelists’ discussion of the evolving landscape of artificial intelligence (AI) focused on recent advancements, legal challenges, regulatory responses, and standard-setting efforts. Dr. Horton began by discussing late-breaking developments in AI technologies, paying special attention to generative AI and the questions it raises around accountability and responsibility. Mr. Ghavi, a lawyer specializing in intellectual property and technology, then provided his perspective on recent AI-related copyright claims. Mr. Satia and Mr. Rossen explored regulatory issues, as well as the Biden administration’s Executive Order on Safe, Secure, and Trustworthy AI. The panelists then turned the discussion to AI compliance programs, examining issues involved in their design, who should implement such programs, and how much they should rely on the end user rather than the upstream developer. The panelists also shared their thoughts on how much – if any – human oversight is needed over AI, as well as the potential risk for algorithmic collusion. The discussion closed with Mr. Rossen noting that company policies are critical for limiting the risk of employees leaking information to AI models, and that there will be a big push over the next year to provide AI literacy work to the public.