Florida appeals court threatens sanctions over 'AI slop' filings

Florida court targets quality, not just accuracy, in AI-assisted legal work

A Florida appeals court has opened a new chapter in judicial scrutiny of artificial intelligence in legal practice, threatening sanctions against an attorney whose filings contained what judges called "AI slop"—lengthy, unfocused submissions filled with incomprehensible prose.

The 4th District Court of Appeal in West Palm Beach ordered family-law attorney Jaclyn Soroka to explain within 10 days why she should not face sanctions for filings the three-judge panel described as containing "frivolous, scattershot arguments" and "gibberish." The court said it would consider referring Soroka to the state bar for disciplinary proceedings.

What distinguishes this case from previous AI-related sanctions is that the judges found no fabricated case citations or factual errors—the hallmark "hallucinations" that have landed other attorneys in trouble. Instead, the panel focused on whether the submissions met basic standards of professional judgment and comprehensibility.

"An AI supercomputer would struggle to find meaning in some of the prose used in this case," wrote Judge Robert Gross in the Wednesday ruling. "Counsel's arguments need to be comprehensible to human readers—the courts, opposing counsel, and the litigants."

Why it matters

This ruling signals that courts are expanding their oversight beyond checking for AI-generated fabrications to evaluating whether attorneys are exercising independent professional judgment when using generative AI tools. The shift suggests that simply ensuring accuracy may not be enough—lawyers must also ensure their AI-assisted work meets baseline standards of clarity and relevance. For legal technology vendors and law firms implementing AI workflows, this creates a new quality-control imperative that goes beyond fact-checking.

A different kind of AI problem

The case emerged from a dissolution of marriage dispute in which Soroka represents the wife and was challenging lower court rulings. Judges Robert Gross, Spencer Levine, and Caroline Shepherd said Soroka's submissions "resemble what we typically see from vexatious pro se litigants" and "appear designed to cause delay."

The panel explicitly distinguished this case from others involving AI hallucinations, stating that while they suspected the filings contained "AI-generated components," the ultimate issue was professional responsibility. "Abusive AI filing threatens the orderly administration of justice," Gross wrote. "Courts must require more from attorneys and not tolerate abusive practice fueled by AI."

The judges emphasized that it doesn't matter whether the work "was AI-generated, AI-assisted, or something else" because attorneys must stand behind any court filings they submit.

Soroka declined an interview but told Reuters she takes the ruling seriously and will respond in court. She would not comment on whether or how she used AI in preparing her filings.

Growing judicial scrutiny

State and federal judges have increasingly sanctioned lawyers over AI use that produces fabricated case citations and legal misstatements. While lawyers are generally not prohibited from using AI tools, they remain ethically bound to ensure the accuracy and quality of court submissions.

This Florida ruling suggests courts are developing more sophisticated frameworks for evaluating AI-assisted legal work, moving beyond simple fact-checking to assess whether submissions reflect the professional judgment expected of licensed attorneys.

These details were first reported by Reuters legal correspondent Mike Scarcella.