Discoverability of AI Prompts and Outputs: Practical Considerations for Counsel
Duty to Preserve, Privilege and Work Product, Relevance and Proportionality, Admissibility, Reverse Prompt Engineering, Proposed FRE 707
About the Course
Introduction
This CLE webinar will offer an overview of how counsel can prepare for the discovery of AI prompts and output from the perspective of both the propounding and the responding party.
Description
A January 2026 Gallup poll concluded that 38% of American workers use AI daily or a few times each week in their jobs and that those in leadership roles (read: people who can bind the company) were more likely to do so. Plausible uses include asking AI for assistance with a legal question to avoid involving counsel, summarizing company documents, asking whether something violates the law or company policies, recording a conversation that later has legal significance, such as demonstrating notice, and comparing competitor pricing, etc.
All the prompts (questions, documents, images, audio/video files, images, computer code, etc.), all the resulting AI generated output (analysis, answers, letters, summaries, transcripts, recommendations, GenAI conversation history, etc.), and sometimes metadata, logs, administrative or audit information are potential targets for discovery in civil actions and criminal or regulatory investigations by or against the company depending on relevance to the facts, claims, and defenses of the case.
Counsel have to be equipped to navigate preservation and collection duties, propound effective discovery requests, argue for or against relevance and proportionality, and deal with privacy and confidentiality issues, as well as custody and control, just as with any other form of ESI. Getting relevant material admitted (or excluded) at trial will require navigating all the same rules of evidence: showing accuracy and reliability, authentication and chain of custody, addressing undue prejudice concerns, privilege and work product, and may require experts to explain how AI systems work. A new Federal Rule of Evidence 707 is under consideration to address some of these issues. Then there is the problem of deep fakes.
Listen as our panel explores the emerging issues and best practices related to the very perplexing, at times overwhelming, and constantly changing questions about the discovery and admission of AI-generated evidence.
Presented By
Mr. Fletcher has over 30 years’ experience as a trial lawyer in complex business litigation. In a span of three months in 2024, he won two Philadelphia jury trials defending cases alleging injuries sustained playing college football. In a span of eight months in 2022, Mr. Fletcher won three jury verdicts for his clients in patent trials held in Texas and Pennsylvania. He has tried numerous jury and non-jury trials to verdict in federal and state courts across the country, and he has led trial teams conducting arbitrations in Stockholm, London, Zurich, and the United States. Mr. Fletcher's trial experience includes torts, patents, contracts, business torts, trade secrets, breach of fiduciary duty, corporate governance, shareholder disputes, trusts, professional liability, and civil rights. He has extensive experience representing technology-focused businesses, including software and systems, chemistry, chemical engineering, and mechanical engineering. On the appellate level, Mr. Fletcher has argued appeals before the U.S. Court of Appeals for the Third, Fourth, and Eleventh Circuits; the Pennsylvania Superior Court; the Pennsylvania Commonwealth Court; and the California Court of Appeal.
Ms. Leibell is a skilled litigator with extensive experience in a variety of complex litigation settings, including class actions and multidistrict litigation. Her experience includes product liability, consumer fraud, insurance, breach of contract, financial services, and intellectual property disputes. Ms. Leibell manages all aspects of discovery in litigation for Fortune 500 companies and works with clients to develop defensible eDiscovery strategies and processes for high stakes matters. She conducts custodian interviews to identify relevant sources of data and information, prepares key witnesses for depositions, takes and defends expert, fact, and 30(b)(6) depositions, crafts discovery requests, and negotiates discovery protocols with opposing counsel as well as handles motions practice addressing discovery issues. Ms. Leibell leverages her many years of trial experience and deep understanding of the Information Law space to provide clients with thoughtful, tailored, and practical legal strategies.
Mr. Weibley focuses his practice on complex commercial litigation, defending Fortune 500, Fortune Global 500, and other corporations in contract and business disputes, intellectual property matters, employment matters, and product liability matters. He represents clients in state and federal trial courts across the country, domestic and international arbitrations, and state and federal appellate courts. In addition to performing all aspects of civil litigation, Mr. Weibley also provides counseling to clients to avoid or best position potential litigation.
-
This 90-minute webinar is eligible in most states for 1.5 CLE credits.
-
Date + Time
- event
Thursday, March 26, 2026
- schedule
1:00 p.m. ET/10:00 a.m. PT
I. What is AI evidence?
II. Applicable discovery rules
A. Duty to preserve AI evidence and preservation letters
B. Best practices for updating legacy discovery requests
C. Analyzing relevancy and proportionality
D. Rule 26(f) conference issues
E. AI prompts as trade secrets, IP, or confidential proprietary information
III. Reverse prompt engineering results as evidence
IV. Planning for admission of AI evidence
A. FRE 901
B. FRE 401, 102, 403
C. FRE 601, 602, 613
D. FRE Article 7 issues
E. Proposed FRE 707
F. Key decisions
1. Tremblay v. OpenAI Inc.
2. In re OpenAI Inc. Copyright Infringement Litigation (MDL, S.D.N.Y.)
3. Motion for a Ruling that Documents the Defendant Generated Through an Artificial Intelligence Tool Are Not Privileged at 7, United States v. Heppner, No. 25-00503-JSR (S.D.N.Y. Feb. 6, 2026), Dkt. No. 22
V. Practical considerations
The panel will review these and other important issues:
- Is discovery of Generative AI evidence, as some have posited, similar to seeking a party's browser history?
- What should a preservation letter say about AI prompts and output, especially if the client would be tempted to ask AI to summarize or evaluate counsel's recommendations?
- What is "reverse prompt engineering," and would the results be admissible as evidence of intent? Or used solely to impeach the AI system?
- Where and how can counsel educate themselves about these issues?
Related Courses
Discoverability of AI Prompts and Outputs: Practical Considerations for Counsel
Thursday, March 26, 2026
1:00 p.m. ET/10:00 a.m. PT
Automatic Renewal and Subscription Laws Update: What’s Coming in 2026 and the Most Important Developments in 2025
Thursday, April 23, 2026
1:00 p.m. ET/10:00 a.m. PT
