Florida Trial Courts Demand Disclosure of AI Use in Pleadings
The two largest judicial districts in Florida will now require lawyers to certify whether artificial intelligence was used in any fashion to create pleadings filed in their courts. Legal research and document drafting, both common uses of generative artificial intelligence, are covered by the certification requirement, as are discovery materials such as deposition summaries. The orders are the latest judicial effort to root out “hallucinations” in court filings by bringing a little Florida sunshine to the use of AI in pretrial practice, though they may also be a catalyst for innovation and more productive uses of these technologies.
The Eleventh Judicial Circuit serves populous Miami-Dade County, with approximately 80 trial judges, making it the largest jurisdiction in Florida. The Seventeenth Judicial Circuit, serving Broward County and based in Fort Lauderdale, is the second largest judicial circuit in Florida. These two recent circuit-wide orders on artificial intelligence will have a significant impact across Florida.
Any attorney or self-represented litigant who uses any generative artificial intelligence tool in the preparation of a pleading, motion, memorandum, response, proposed order, or other court record, must disclose such use on the face of the filing.
Pleadings filed in the Eleventh Judicial Circuit must now carry the following AI certification:
The filing shall include a statement substantially in the following form: “Generative artificial intelligence was used in the preparation of this filing. The undersigned certifies that all factual assertions, legal authority, and citations have been independently reviewed and verified for accuracy and accepts full responsibility for the contents of this filing.”
Eleventh Circuit Order 26-04, published Jan. 15, 2026, is intended to root out AI-generated “hallucinations” in pleadings. Trial judges in Florida, and elsewhere, have increasingly expressed frustration with errant citations to facts and law in pleadings and briefs. Chief Judge Ariana Fajardo Orshan noted in her order, “the Court should not have to parse case citations and parentheticals to discern whether cases exist.”
Going forward, “[a]ny attorney or self-represented litigant who uses any generative artificial intelligence tool in the preparation of a pleading, motion, memorandum, response, proposed order, or other court record, must disclose such use on the face of the filing.”
In the Seventeenth Judicial Circuit (Broward County), all pleadings must now carry the following certification:
The undersigned hereby certifies that generative artificial intelligence was used to prepare this [TITLE OF DOCUMENT BEING FILED]. The undersigned has independently verified the accuracy of every citation to the law and/or the record, and the accuracy of any language drafted by generative artificial intelligence, including quotations, citations, paraphrased assertions, facts, and legal analysis.
Seventeenth Circuit Administrative Order 2026-03-Gen, issued Jan. 26, 2026, by Chief Judge Carol-Lisa Phillips, explicitly approves the use of artificial intelligence in pretrial discovery and pleadings, but cautions that the technology must be used within professional ethical obligations: client confidentiality, candor to the tribunal, diligence, compliance with court rules of procedure and with other Florida rules regulating the practice of law. “Changes in the law and its practice exist including benefits and risks associated with the technology relevant to courtroom proceedings,” the order states. “If a person uses artificial intelligence tools or not, the understanding of its uses and challenges is essential to the preparation of any filing or presentation.”
Violations of these new certification requirements can be punished by monetary sanctions and referral to the Florida Bar for disciplinary proceedings.
Arguably, the representations explicitly demanded by these two Florida courts are already required by existing court rules on the significance of an attorney’s signature on pleadings. In the federal system, for example, Rule 11 of the Federal Rules of Civil Procedure provides that attorneys, when signing pleadings filed in federal court, certify that assertions of fact and statements of law have evidentiary support.
There is no uniform nationwide requirement for lawyers to disclose the use of artificial intelligence in preparing court pleadings. Instead, AI disclosure requirements vary significantly by jurisdiction, with some courts requiring mandatory disclosure, some banning AI use entirely, and many having no specific requirements at all. In view of the current patchwork compliance environment, the default practice in many firms across the country is to include an AI certification in all relevant documents filed with any court.
The Problem of Secret Task Automation
AI certification requirements, in Florida and elsewhere, are plainly intended to promote accountability, fairness, and public confidence in the judicial process. Beyond these immediate objectives, AI disclosure requirements might also serve to encourage innovation and may actually improve the way artificial intelligence is used in the practice of law.
In his 2024 book Co-Intelligence: Living and Working with AI, Wharton School professor and AI researcher Ethan Mollick identified a workplace phenomenon he called “secret task automation.” Many workers, he noted, were surreptitiously using artificial intelligence to automate parts of their jobs without informing their organizations.
Mollick identified several reasons why AI was being used in secret. First, some workers were coping with organizational bans on the use of AI in their jobs. Second, these secret AI users believed that people judge AI-created materials differently — and worse — than human-created content. (Mollick cites research indicating evidentiary support for this belief.) Third, fears of being replaced lead some workers to keep quiet about AI-enabled efficiencies.
In Mollick’s view, “secret task automation” among lower-level employees deprives their organizations of the full benefits of artificial intelligence. Few employees today were hired based on their AI skills; thus, AI expertise could be anywhere in the organization. Employees using AI most productively could very well be found in an organization’s lowest levels, not in management. Mollick encouraged management officials to take steps to decrease fears associated with revealing AI use, such as guaranteeing that AI use will not result in job loss, offering financial incentives for AI use, and promising more interesting work with the time gained via AI efficiencies.
When seen through this lens, court-ordered AI disclosure requirements may well be a boon and not a burden to lawyers and their firms. Mandatory disclosure of the use of AI in court pleadings will stimulate firm-wide discussions around the responsible use of these technologies for client work. Law firm managers will more readily be able to meet their obligation under Model Rule of Professional Conduct 5.1 (Responsibilities of Partners, Managers, and Supervisory Lawyers) to take reasonable measures to ensure ethical compliance by all firm personnel and outside vendors.
Finally, with a new, court-ordered window into how and where artificial intelligence is being used by litigators, senior law firm personnel will be more likely to think strategically about the use of these tools and to capture their most productive and effective uses in litigation — all to the client’s benefit.