When the AI Acts: A Case Study in Agentic AI, Legal Ethics, and Client Risk

Live · 2 CLE credits · November 17, 2026, 12:00 PM ET

AI agents no longer just answer questions. They draft, file, send email, and process refunds. Apply the Model Rules when your agent acts, build AI use policies with approval gates and human review, and leave with deployment and contracting checklists.

When the AI Acts, the Lawyer Still Answers

AI agents no longer stop at answering questions. They draft and file, send email, process refunds, and enter into transactions with limited human involvement. That shift raises new questions for lawyers who use these tools and for lawyers who advise the companies that deploy them.

File a brief your research agent drafted, and fabricated citations invite sanctions under Rule 11 and 28 U.S.C. § 1927. Give an agent inbox access, and privileged comments can leave the firm, testing Rule 1.6(c) and FRE 502. Let a client's agent act beyond its authority, and apparent authority under Moffatt v. Air Canada becomes the client's problem. Launch without a review, and the EU AI Act, state AI laws, and FTC enforcement against AI washing follow.

The program follows one hypothetical company, Meridian Health Retail, Inc., and its outside counsel through eight escalating scenes. Each scene closes with a discussion question and an audience poll. You leave with a framework for agentic AI use policies built on permissions, logging, approval gates, and human review. In-house counsel get a deployment checklist, and outside counsel get an advising and contracting checklist. Every attendee receives the Meridian fact pattern before the program.

Key topics

Agenda

Faculty

State accreditation