AI-Assisted Pro Se Litigation Requires Early In-House Scrutiny

Growing use of generative artificial intelligence by pro se parties in consumer finance litigation is compelling in-house counsel to consider seeking early guidance from trial courts.

More self‑represented filings have been arriving polished, requiring defendants to spend resources verifying whether the cited authorities exist and whether they support what the filer claims. That verification burden is now a recurring theme in both reporting and judicial commentary, especially as courts confront filings with the hallmarks of AI “hallucinations.”

This is an expensive trend, especially for consumer‑finance defendants. The result is immediate: more filings and more hearings. Because pro se litigants aren’t paying counsel, litigation costs can become asymmetrical and unnecessarily increase defense spending.

Courts increasingly are signaling that they want accountability that fits within existing doctrines, including disclosure of AI use and human verification of citations and factual assertions. Recent case law and administrative orders show that AI governance is expanding into confidentiality, protective orders, and discovery disputes about prompts and outputs.

Here are some ways that legal departments can anticipate those types of issues earlier rather than later:

Seek a comprehensive pro se AI order early that is routine and venue-aware. At the outset of a newly filed pro se matter, in‑house counsel should direct litigation counsel to seek a case‑specific scheduling order that addresses generative AI use in filings—before the docket fills with dubious pleadings.

Florida’s 11th and 17th judicial circuits, for example, issued administrative orders requiring self‑represented litigants to disclose generative AI use and certify independent verification of citations and factual assertions, with sanctions for noncompliance. Those orders matter because they reflect mainstream court administration treating AI disclosure and verification as a procedural baseline, not because they’re unique.

Addressing AI‑hallmark issues in a pro se filing, Florida’s Sixth District Court of Appeal emphasized the obligation to cite‑check and required remedial steps while issuing a show‑cause order. The US Court of Appeals for the Seventh Circuit’s discussion in Jones v. Kanakee County Sheriff’s Department similarly underscores that even pro se filings are representations to the court and require reasonable care to avoid misrepresentations.

Counsel should treat the AI order as a template‑driven, low‑cost early motion that becomes routine in pro se matters. The order should require filing‑by‑filing disclosure of AI assistance, human verification of citations, quotations, and key factual assertions, plus a clear sanction notice.

Build prompts and outputs into discovery. As AI use expands, prompts, outputs, chat histories, and AI‑generated drafts can become relevant to the provenance of filings, the integrity of citations and quotations, and the good‑faith basis of what is submitted to the court.

The key is to make this routine, targeted, and proportional. In‑house counsel should:

• Direct litigation counsel to incorporate AI usage questions into standard discovery templates
• Identify what tools were used and whether histories were deleted or disabled
• Request prompts and outputs tied to the drafting of pleadings or motions and to the factual narratives on which the pro se party relies

Courts may treat some AI‑assisted litigation preparation materials as protected work product in the pro se context. In Morgan v. V2X, Inc., the court addressed how work‑product principles apply to a pro se litigant’s AI use while compelling tool identity disclosure and amending the protective order to address AI risk. The practical implication isn’t that prompts and outputs are always protected or always discoverable; it’s that courts will take a fact‑specific approach.

Ensure preservation letters explicitly cover AI. Consumer AI tools allow users to delete histories, and self‑represented litigants may not treat prompts and outputs as litigation records. A preservation letter that explicitly includes AI prompts, outputs, drafts, and history settings reduces the risk of negligence and clarifies preservation duties early.

Preservation language should be part of the first wave correspondence in pro se cases, not an optional add‑on after suspicious citations appear. It’s reasonable to treat AI artifacts as ordinary electronically stored information that must be preserved when relevant and proportional. But courts still may recognize work‑product protections in some contexts, particularly for litigation preparation.

Prevent predictable confidentiality risks. In consumer‑finance litigation, confidentiality isn’t always the primary driver of cost. Confidentiality becomes significant the moment AI intersects with sensitive borrower information and discovery material.

Broward County, Fla.’s administrative order includes “confidentiality and authenticity” provisions and applies to both attorneys and pro se litigants. Morgan illustrates how quickly a case can pivot into a protective‑order dispute when a pro se party uses AI in connection with confidential discovery that causes tool identity disclosure.

Counsel shouldn’t wait to “clean up” a disclosure after the fact. Instruct litigation counsel at the outset to incorporate AI‑specific protective‑order guardrails designed to prevent upload of confidential borrower information into AI tools.

Design proactively. Generative AI has produced plausible-looking pro se filings and increased motion practice, which translates into an assortment of motions to quash, motions to vacate and motions for reconsideration.

With each new motion, commercial defendants and their counsel are compelled to file responses, and additional briefing, in opposition. Courts are responding with disclosure and verification expectations that apply to self‑represented parties, while addressing AI as a confidentiality and protective‑order issue when it intersects with sensitive information.

Legal departments should instruct litigation counsel to fully incorporate these evolving developments into a more routine litigation management approach that spans the lifecycle of litigation in the generative AI era, while retaining concepts and structures that are familiar to trial courts.

Clear and well-developed preservation instructions should be a pre-cursor to all relief that a commercial litigant might seek. Even in jurisdictions where a trial court has an existing standing order, litigation counsel should seek clarifying orders when the standing order may not be fully developed. Requests around prompts, outputs and tools should be part of the discovery template.

Legal departments should assume that they are the first line of defense in protecting confidential information.

An immaterial amount of this content was drafted by generative artificial intelligence.

This article does not necessarily reflect the opinion of Bloomberg Industry Group, Inc., the publisher of Bloomberg Law, Bloomberg Tax, and Bloomberg Government, or its owners.

Author Information

Lee Barrett is senior vice president and associate general counsel at Planet Home Lending.

Interested in writing? Review our author guidelines, and submit pitches to Insights@bloombergindustry.com.