Gleason v. Marcus: Florida Appeals Court Warns Pro Se Litigant Over AI-Hallucinated Election Law Brief | Advocate Prakhar

Case at a Glance

Background

Christopher Gleason, proceeding pro se, challenged a circuit court order dismissing his complaint against the Pinellas County Supervisor of Elections. The substance of his election-related claims was found to lack merit, and the district court of appeal affirmed the dismissal without further comment on the underlying issues. What drew the court’s attention — and prompted a written opinion — was not the merits but the quality of Gleason’s appellate briefs.

A review of the briefs found multiple citations to cases that simply do not exist, as well as citations to real cases attributed with legal propositions that appear nowhere in those opinions. The pattern is the hallmark of generative AI that has been used to produce legal arguments without any subsequent verification against actual legal databases. The court had previously addressed this problem in Russell v. Mells (2025) involving an attorney, and now encountered it in a pro se appeal.

The AI Issue

The Florida Second DCA’s concern, as expressed in this case and in its cross-citation to the Fourth DCA’s Gouveia v. Meridian Financial Investments concurrence, is systemic: AI tools are “very good at sounding right, but less adept at being right,” and “blind reliance” on them is unacceptable. The court is not prohibiting AI use; it is demanding verification. The failure here — as in so many cases — is not using AI, but using AI as a substitute for the simple acts of checking whether a case exists and whether it stands for the cited proposition.

What the Court Decided

• Final order of dismissal affirmed on the merits without further analysis.
• Court wrote separately to address Gleason’s AI-hallucinated briefs.
• Gleason formally notified that future filings of the same nature will not receive favourable treatment.
• All litigants — pro se and represented — reminded that they must ensure the accuracy of their filings under Florida Rule of Appellate Procedure 9.045(d) and Florida Rule of General Practice and Judicial Administration 2.515(d)(2).
• Court cited line of Florida AI-hallucination cases showing an escalating jurisprudential trajectory: from warning (this case) to sanctions (prior and subsequent cases).

“This technology is very good at sounding right, but less adept at being right, and blind reliance on it is unwise, unacceptable, and sanctionable.”

— Florida 4th DCA, Gouveia v. Meridian Financial Investments (2026), cited approvingly in Gleason v. Marcus

The India Angle

Indian Law Equivalent

Election disputes in India are governed by the Representation of the People Act, 1951, and the rules made thereunder. Election petitions are heard by High Courts under Chapter IV of Part VI of the Act. The obligation of candour to the court in election petition proceedings is particularly high because such petitions affect democratic rights. Under the Election Laws (Amendment) Act and constitutional jurisprudence, courts have consistently held that election disputes must be decided on verified facts and authentic legal authority. AI-hallucinated citations in an election petition would engage the court’s contempt powers and could lead to dismissal of the petition as an abuse of process.

Bar Council Rules

The BCI Rule on misstatement of law (Rule 49) applies with equal force in election law proceedings. Given the public interest dimension of election disputes, a court finding that an advocate filed AI-hallucinated citations in an election petition is likely to treat it as more serious misconduct than in ordinary commercial litigation. The Supreme Court and High Courts have wide suo motu powers to address conduct that undermines the integrity of election jurisprudence.

Practical Advice for Indian Advocates

• When an appellate court issues a formal warning about your filing practices, treat it as a final warning — a subsequent violation in the same court, or in another court that cites the warning, will likely result in punitive sanctions.
• Florida courts have explicitly built a citation trail from Gutierrez v. Gutierrez (sanctions for pro se) through Russell v. Mells (sanctions for attorney) to this case — Indian courts can be expected to build similar precedential chains once the first major AI-sanction ruling is issued by the Supreme Court or a High Court.
• Pro se litigants in India who use AI for legal research should understand that being unrepresented does not lower the accuracy standard — accuracy of citations is a baseline obligation regardless of legal training.

Quick Takeaways

• A formal court warning about AI hallucinations is a precursor to sanctions — it establishes notice and eliminates the defence of ignorance in future proceedings.
• Florida’s appellate courts have developed a consistent multi-court jurisprudence on AI hallucinations spanning at least five decisions and multiple districts.
• The verification obligation — cite-check existence, cite-check accuracy of proposition — is simple to satisfy but requires a disciplined post-drafting step that AI cannot perform for you.

Deep Dive: Florida’s Escalating AI-Hallucination Jurisprudence — From Gutierrez to Gleason

Florida’s appellate courts have developed what is now a discernible and escalating body of AI-hallucination jurisprudence. The trajectory runs from Gutierrez v. Gutierrez (3d DCA 2024, sanctions for pro se litigant), through Goya v. Hayashida (4th DCA 2025, sanctions and warning that fake opinions abuse the adversary system), Takefman v. Pickleball Club (3d DCA 2025, attorney AI misuse), Russell v. Mells (2d DCA 2025, attorney referral and sanctions), Hessert v. Hessert (6th DCA March 2026, detailed warning with guidance to opposing counsel), Gouveia v. Meridian Financial (4th DCA March 2026, judicial concurrence calling AI hallucinations “unacceptable, and sanctionable”), and now Gleason v. Marcus (2d DCA May 2026). At least seven distinct Florida appellate rulings in roughly 18 months — each building on the last.

What this trajectory reveals is that Florida courts are deliberately building a body of precedent that will support increasingly severe consequences. The first cases established that AI hallucinations violate professional rules. The next cases imposed monetary sanctions. Then came referrals to the Florida Bar. The Hessert case established that opposing counsel have an obligation to point out AI errors in responsive filings. Gleason issues a formal pre-sanctions warning, and cross-references the full line of cases to ensure that no future litigant can claim ignorance of the stakes.

The practical message for Indian advocates is that a similar trajectory will likely unfold in India once the first major AI-hallucination ruling is issued by a superior court. Indian courts — particularly the Supreme Court and High Courts — tend to build precedential chains in systemic-misconduct areas: from isolated rulings to guidelines (as happened with Vishaka, with suo motu PIL practice, and with standards for expert witnesses). When the first Indian AI-hallucination ruling comes, it will likely reference the global jurisprudence — including Florida’s — and use it as scaffolding for what will quickly become a comprehensive Indian standard.

The strategic implication: advocates who establish best practices now, before Indian courts have issued formal guidance, position themselves ahead of the curve. Those who wait for an Indian equivalent of Mata v. Avianca before treating AI verification as a professional obligation will find themselves responding to a jurisprudence that was entirely predictable and that their peers had already internalised.