Protective Orders That Bar Public AI Tools: The Producing Party’s Motion and Clause

Patrick Fanning
David A. Berkley
Kristina Krull
Patrick Fanning | Lathrop GPM LLP
David A. Berkley | Womble Bond Dickinson (US) LLP
Kristina Krull | Womble Bond Dickinson (US) LLP

Live Video-Broadcast: December 14, 2026

2 hour CLE

Tuition: $195.00
Subscribe to Federal Bar Association CLE Pass...
Co-Sponsored by myLawCLE
Get this course, plus over 1,000+ of live webinars.
Learn More
Training 5 or more people?

Sign-up for a law firm subscription plan and each attorney in the firm receives free access to all CLE Programs

Program Summary

 

A Standard Protective Order Will Not Keep Your Production Out of a Public AI Tool

Courts are splitting on whether a protective order can restrict an opponent’s use of public generative AI tools on confidential discovery. In Jefferies v. Harcros the court entered such limits. In Litton v. Roblox the court declined to add them to a model order. Whether your client’s production gets that protection depends on your record and your clause.

The stakes sit with the producing party. Raise the issue at the Rule 26(f) conference and the restriction is on the table before production. Wait, and you are moving to amend after the documents are out. Bring a thin declaration and the objections that lost in Jefferies win. Leave AI Tool, closed environment, train, retain, and outputs undefined and the clause has no edges to enforce. Restrict tools your own firm and vendors use and the restriction backfires on the producing party.

This two-session program gives producing-party counsel the motion and the clause. Attendees leave with a sequencing decision tree, a declaration checklist, and practical guidance on which record items actually move a magistrate judge. They gain annotated alternative clauses from Jefferies, Orechovesky, Morgan, Moher, ByHeart, and Litton. A low-cost fallback clause and a reciprocity audit worksheet round out the toolkit.

Key topics to be discussed:

  • The Sequencing Call
    Decide between the Rule 26(f) conference, a motion to amend after production, or a held enforcement request.
  • The Good-Cause Record
    Build on the Jefferies record items that carried good cause and tailor the regulatory-exposure and technical declarations to patient charts, personnel files, payroll data, and medical records.
  • Model-Order Districts
    Bring the stand-alone motion after Litton and fit the proposed order to the court’s test with tool identity as an enforcement term.
  • Scope and Tool Standards
    Choose between all Discovery Materials and Confidential material only, and read the competing tool standards side by side.
  • Drafting the Clause
    Draft from neutral functions after Morgan’s rewrite and within the model-order limit, with AI Tool, closed environment, train, retain, and outputs defined.
  • Flow-Down and Reciprocity
    Flow the restriction down to vendors and experts, add compliance hooks and the deletion certification, and run the reciprocity audit so the clause does not backfire, with a low-cost fallback clause in reserve.

This course is co-sponsored with myLawCLE.

Date / Time: December 14, 2026

  • 12:00 pm – 2:10 pm Eastern
  • 11:00 am – 1:10 pm Central
  • 10:00 am – 12:10 pm Mountain
  • 9:00 am – 11:10 am Pacific

Closed-captioning available

Speakers

Patrick (Pat) Fanning, Partner | Lathrop GPM LLP

Pat Fanning is a Partner in the Kansas City office of Lathrop GPM LLP, where he tries complex business disputes and toxic tort matters that sit at the intersection of technology, science, and contract law. A trial attorney with more than twenty-five years of experience, he has taken numerous jury trials and commercial arbitration hearings to verdict and award, and for several years he has defended a Fortune 500 information technology company in software disputes with other businesses. He is also one of the litigation bar’s leading voices on artificial intelligence in discovery: he is the author of “A New Approach for Protection of ‘Nonconfidential’ Documents Produced in the Age of AI” (The Legal Intelligencer, June 2026), the very problem this program addresses.

  • Education & Credentials

Mr. Fanning earned his J.D. from the University of Missouri-Columbia School of Law in 1998 and his B.A. from the University of Missouri-Columbia in 1995. He is admitted to practice in Missouri and Kansas, and before the U.S. Courts of Appeals for the Second, Sixth, Seventh, and Eighth Circuits, the U.S. District Court for the Western District of Missouri, and the U.S. District Court for the District of Kansas.

  • Recognition & Leadership

Mr. Fanning holds Martindale-Hubbell’s highest peer rating, AV Preeminent. He has been recognized by The Legal 500 United States (2014) and named to Missouri & Kansas Super Lawyers (2017, 2020–2021), after earlier selections as a Super Lawyers Rising Star (2008–2010, 2013–2014). His views on where the technology is taking the profession were featured in Opus 2’s “The Future of AI in Litigation: Predictions and Advice from Legal Industry Experts” (January 2026), and he has been quoted on emerging litigation risks by DataCenter Knowledge and the Kansas City Business Journal (2026).

  • Professional Involvement

Mr. Fanning is a regular presenter at Lathrop GPM’s annual State of Litigation program, where he co-presented “You Can’t Spell ‘Inevitable’ Without A.I.” (2024) and “When Goliath Borrows David’s Slingshot: Big Business and its Offensive Use of State Consumer Protection Acts” (2023), and delivered the welcome remarks in both years. In the community, he has served on the Board of Directors of the FIRE Foundation (2015–2021), serves on the Faith Formation Committee of Visitation Catholic Church, coaches cross-country at Visitation Catholic School, and is a graduate of the Greater Kansas City Chamber of Commerce Centurions Leadership Program.

  • Experience

A Kansas City native and second-generation trial attorney, Mr. Fanning began his career at Lathrop & Gage LLP, where he became a partner in 2007. He then spent roughly eight years as a partner at a Kansas City litigation firm and was a founding partner of PEAK Litigation LLP, a commercial and tort litigation boutique, before rejoining Lathrop GPM in 2021. His results include a complete defense judgment on business tort claims seeking more than $132 million after a two-week trial conducted by Zoom; a preliminary injunction, won after a multi-day hearing, that kept a client’s vendor software running when the vendor threatened to cut it off; an AAA arbitration award after a week-long hearing against a party trying to walk away from an implemented software contract; the resolution of multiple arbitrations and lawsuits arising from a client’s $1 billion acquisition; and jury verdicts in federal and state court, including a lengthy federal trial in a dispute between a Kansas corporation and a state-owned factory in the People’s Republic of China. Having litigated high-stakes technology disputes where the documents produced are the client’s crown jewels, and having written on how to protect productions from AI tools, Mr. Fanning is uniquely positioned to walk producing-party counsel through the motion and the clause.

 

David A. Berkley, Partner | Womble Bond Dickinson (US) LLP

David A. Berkley is a Partner in the Irvine, California office of Womble Bond Dickinson (US) LLP, where he is a commercial litigator with deep experience in financial services and labor and employment disputes. He has litigated hundreds of matters and is involved in every phase of a case, from the initial filings through closing argument. Mr. Berkley is also the co-author of “The Need to Address Open-Source Artificial Intelligence in Bespoke or Model Litigation Protective Orders” (Womble Bond Dickinson, April 2026), which analyzes the gap in courts’ model protective orders and the emerging case law, including Morgan v. V2X, Inc. (D. Colo. 2026), restricting the input of confidential discovery material into AI platforms. That is the precise subject of this program.

  • Education & Credentials

Mr. Berkley earned his J.D. from Loyola Law School, Los Angeles, his M.B.A. from Loyola Marymount University, and his B.A. from the University of Pennsylvania. He is admitted to practice in California and before the U.S. Court of Appeals for the Ninth Circuit.

  • Recognition & Leadership

Mr. Berkley has been recognized in The Best Lawyers in America for Commercial Litigation (2025–2026) and was selected to the Southern California Super Lawyers Rising Stars list for three consecutive years (2015–2017). His combined legal and business training informs the practical, cost-conscious advice he gives clients on managing litigation risk.

  • Professional Involvement

Mr. Berkley writes and speaks on litigation practice. In addition to his 2026 analysis of AI and protective orders, co-authored with his colleague Scott J. Hyman, he served as faculty for the seminar “Litigating Civil Cases in California From Start to Finish” (2023). He practices within Womble Bond Dickinson’s Litigation and Dispute Resolution and Consumer Finance Disputes and Litigation groups and its Financial Services sector team.

  • Experience

Mr. Berkley has litigated hundreds of matters involving banking operations and mortgage-based actions, as well as claims under the Fair Credit Reporting Act, the Telephone Consumer Protection Act, and the Fair Debt Collection Practices Act. His employment practice includes the defense of sexual harassment, discrimination, and wrongful termination claims, and he has trial preparation experience in complex business litigation, commercial real estate disputes, and insurance and personal injury claims. Representing financial institutions whose productions are full of confidential customer and business information, and having studied how courts are now treating AI platforms under existing protective orders, Mr. Berkley is uniquely positioned to show producing-party counsel how to build the record for the motion and draft a clause a court will enter.

 

Kristina Krull, Associate | Womble Bond Dickinson (US) LLP

Kristina Krull is an Associate in the San Francisco office of Womble Bond Dickinson (US) LLP and a member of the firm’s Business Litigation Group. Her practice centers on financial services disputes and litigation, with a particular focus on data privacy law. She defends banks and other financial institutions in state and federal courts and in arbitration, in matters where the confidentiality of customer information and the handling of sensitive records are front and center. That daily work with privacy statutes and confidential financial data gives her a practical, document-level view of what a producing party stands to lose when discovery material is fed into a public AI tool.

  • Education & Credentials

Ms. Krull earned her J.D. from Santa Clara University School of Law in 2021 and her B.A. in Communication Studies from the University of Maryland in 2017. She is admitted to practice in California.

  • Recognition & Leadership

At Santa Clara Law, Ms. Krull served as Program Director of the Honors Moot Court External Board and competed in the American Bar Association Negotiation Competition, early training in the advocacy and negotiation skills that drive contested protective-order practice. At the University of Maryland she was a four-year student-athlete on the women’s tennis team.

  • Professional Involvement

Ms. Krull practices within Womble Bond Dickinson’s Business Litigation Group and its Financial Services sector team, working alongside the firm’s consumer finance and privacy litigators on matters for banks and financial institutions across California. Her work regularly requires her to apply California’s privacy laws in active litigation, the same body of law that shapes how confidential information must be protected once it is produced.

  • Experience

Ms. Krull has litigated auto finance, wire fraud, faithless bookkeeper, and mortgage servicing disputes in state and federal courts and in arbitration. She has defended banks in a range of disputes involving fraud and forgery, including wire fraud, and legal order processing; defended banks in cases brought under California privacy laws in both state and federal court; and litigated mortgage servicing disputes involving predatory lending and wrongful foreclosure claims. Because her clients’ productions routinely contain nonpublic personal and financial information, Ms. Krull is well positioned to explain how producing-party counsel should frame the confidentiality interests at stake and draft protective-order language that keeps that information out of public AI tools.

Agenda

SESSION 1 – Building the Good-Cause Record to Restrict an Opponent’s Public AI Use | Kristina Krull & David A. Berkley | 12:00pm – 1:00pm

This session equips federal civil litigators who represent producing parties, from either side of the caption, to build the good-cause record needed to restrict an opponent’s use of public generative AI tools on confidential discovery. Drawing on Jefferies v. Harcros, where the court entered such limits, and Litton v. Roblox, where the court declined to add them to a model order, the session covers the timing decision between the Rule 26(f) conference and a motion to amend after production; the record items that carried good cause and the objections that lost; declarations tailored to the client’s fact pattern (patient charts, personnel files, payroll data, medical records); the stand-alone motion in model-order districts; and tool identity as an enforcement term under Morgan v. V2X. Attorneys completing this session will leave with a sequencing decision tree, a declaration checklist, and practical guidance on which record items actually move a magistrate judge.

BREAK | 1:00pm – 1:10pm

SESSION 2 – Drafting the AI-Use Clause in a Producing Party’s Protective Order | Patrick (Pat) Fanning | 1:10pm – 2:10pm

Building on Session 1’s focus on winning the restriction, this session turns to drafting the AI-use paragraph itself, working from the clauses courts have actually entered in Jefferies, Orechovesky, Morgan, Moher, ByHeart, and Litton. The session walks through the scope decision (all Discovery Materials or Confidential material only); the competing tool standards read side by side; drafting from neutral functions after Morgan’s rewrite and within the model-order limit; defined terms such as AI Tool, closed environment, train, retain, and outputs; flow-down to vendors and experts, compliance hooks, and the deletion certification; and a reciprocity audit of the drafting firm’s own tools and vendors so the restriction does not backfire on the producing party. Attendees will leave with annotated alternative clauses, a low-cost fallback clause, and an audit worksheet they can apply in their next case.

Credits

Alaska

Approved for CLE Credits
2 General

Our programs are CLE-eligible through Alaska’s recognition of multi-jurisdictional reciprocity.
Alabama

Pending CLE Approval
2 General

Arkansas

Approved for CLE Credits
2 General

Arizona

Approved for CLE Credits
2 General

California

Approved for CLE Credits
2 General

Colorado

Pending CLE Approval
2 General

Connecticut

Approved for CLE Credits
2 General

District of Columbia

No MCLE Required
2 CLE Hour(s)

Delaware

Pending CLE Approval
2 General

Florida

Pending CLE Approval
2 General

Georgia

Pending CLE Approval
2 General

Hawaii

Approved for CLE Credits
2 General

Iowa

Pending CLE Approval
2 General

Idaho

Pending CLE Approval
2 General

Illinois

Pending CLE Approval
2 General

Indiana

Pending CLE Approval
2 General

Kansas

Pending CLE Approval
2 Substantive

Kentucky

Pending CLE Approval
2 General

Louisiana

Pending CLE Approval
2 General

Massachusetts

No MCLE Required
2 CLE Hour(s)

Maryland

No MCLE Required
2 CLE Hour(s)

Maine

Pending CLE Approval
2 General

Michigan

No MCLE Required
2 CLE Hour(s)

Minnesota

Pending CLE Approval
2 General

Missouri

Approved for CLE Credits
2.4 General

Mississippi

Pending CLE Approval
2 General

Montana

Pending CLE Approval
2 General

North Carolina

Pending CLE Approval
2 General

North Dakota

Approved for CLE Credits
2 General

Our programs are CLE-eligible through North Dakota’s recognition of multi-jurisdictional reciprocity. Section 1, Policy 1.14
Nebraska

Pending CLE Approval
2 General

myLawCLE reports attendance to Nebraska on each attorney’s behalf for all programs. Please do not self-report.
New Hampshire

Approved for CLE Credits
120 General minutes

As of July 1, 2014, the NHMCLE Board no longer provides pre- or post-approval of courses. Attendees must self-determine whether a program is eligible for credit, and self-report their attendance online at www.nhbar.org, based on qualification provisions of Rule 53.
New Jersey

Approved for CLE Credits
2 General

Our programs are CLE-eligible through New Jersey’s recognition of multi-jurisdictional reciprocity, except for the courses required under BCLE Reg. 201:2
New Mexico

Approved for CLE Credits
2 General

Nevada

Pending CLE Approval
2 General

New York

Approved for CLE Credits
2 General

Our programs are CLE-eligible through New York’s Approved Jurisdiction Group “B”.
Ohio

Pending CLE Approval
2 General

Oklahoma

Pending CLE Approval
2.5 General

Oregon

Pending CLE Approval
2 General

Pennsylvania

Approved for CLE Credits
2 General

Rhode Island

Pending CLE Approval
2.5 General

South Carolina

Pending CLE Approval
2 General

South Dakota

No MCLE Required
2 CLE Hour(s)

Tennessee

Pending CLE Approval
2 General

Texas

Approved for CLE Credits
2 General

Utah

Pending CLE Approval
2 General

Virginia

Not Eligible
2 General Hours

Vermont

Approved for CLE Credits
2 General

Washington

Approved via Attorney Submission
2 Law & Legal Hours

Receive CLE credit in Washington via attorney submission.
Wisconsin

Pending CLE Approval
2 General

West Virginia

Pending CLE Approval
2.4 General

Wyoming

Pending CLE Approval
2 General

More CLE Webinars
Upcoming CLE Webinars
AI-Augmented Legal Workflows: A Map to Success
AI-Augmented Legal Workflows: A Map to Success Thu, November 19, 2026
On-Demand
Live Replay
Using Adobe Acrobat’s AI Assistant for Legal
Using Adobe Acrobat’s AI Assistant for Legal Fri, November 20, 2026
Live Webcast
Obtaining and Using Social Media Evidence (2026 Edition)
Obtaining and Using Social Media Evidence (2026 Edition) Tue, November 24, 2026
On-Demand
Live Replay