A federal judge in Alabama has done something rare: documented, in exhaustive and public detail, what professional incompetence looks like when it finally meets its reckoning. The result is a 12,000-word opinion that reads less like a legal order and more like a forensic autopsy of a legal career.
The case is Heimkes v. Fairhope Motorcoach Resort Condo. Owners Association, decided by Judge Terry Moorer of the Southern District of Alabama. At its center is attorney Franklin Hollis Eaton, Jr., who has been ordered to pay $55,597 in sanctions, recommended for a finding of incompetence to practice law, and required to attach this opinion — unfiled under seal — to every active case he currently handles and every new matter he takes on for the next twelve months. The Alabama State Bar has been notified. So have the chief judges of all three federal districts in the state. The opinion will be published in the Federal Supplement, meaning it becomes permanent, searchable, citeable law.
A Pattern, Not a Mistake
Reason's coverage focuses on the AI-hallucination angle — the fabricated case citations that triggered the formal sanctions proceedings — but the opinion itself is at pains to situate those fabrications within a longer record. Judge Moorer writes that "this case is not just about Mr. Eaton's most recent misstatements of law, but rather about a pattern of conduct throughout this case that raises significant concerns about his competency to practice law." The bogus citations, in other words, were not the story. They were the final straw.
The pattern Moorer documents stretches back to the earliest stages of litigation. Eaton failed to include a jury demand in either of his complaints. The court laid out a clear path for him to correct this and he did not follow it. Co-counsel repeatedly sought to withdraw. Correspondence entered into the record suggested Eaton had filed motions over his co-counsel's explicit objections and had placed their signatures on documents without their consent. He failed to participate in drafting the joint pretrial document as required by standing court order. At the pretrial conference, he told the judge his case-in-chief would take two days. It took twelve.
The trial itself draws some of the opinion's most withering prose. Moorer describes examinations as "plodding, rambling, unfocused" and notes that he was "loathe" to impose time limits on an attorney's witness questioning — but ultimately did so. "Several minutes would pass between the witnesses answering a question and Mr. Eaton asking the next question." The judge concluded that Eaton appeared to be calling witnesses not because they were necessary to his case, but because they happened to be present in the gallery — essentially using the federal trial as a discovery vehicle for parallel state litigation.
The Fabrications
After the bench trial concluded, Eaton filed a response opposing the defendant's motion for directed verdict. The court noticed what it described as "blatant discrepancies" in his citations and began an independent review. Defense counsel moved faster, filing a motion to strike that identified cases that did not exist, citations that led to the wrong case names, and cases that, even when real, had nothing to do with the legal propositions they were cited for.
The court's own search confirmed every allegation. As Reason reports, the piece details that "the cases either do not exist, citations did not go to the referenced names, and/or the cases did not relate to the issues at hand." This is the fingerprint of a large language model fabricating legal authority — a phenomenon well-documented since 2023, when the first wave of AI-hallucination sanctions cases began appearing in federal courts. An AI tool, when asked to find supporting caselaw, will often produce plausible-sounding citations to cases that do not exist. An attorney who submits those citations without verification has, in effect, lied to the court.
Judge Moorer issued two separate show cause orders, giving Eaton repeated opportunities to explain, correct, and account for the errors. He never fully did. He showed up to the first show cause hearing entirely unprepared. He requested continuances. He offered explanations that the court found incomplete or implausible. He never addressed all of the court's identified misstatements, even after they were laid out for him explicitly.
"The insertion of bogus citations and misrepresentation of authorities is not a mere typographical error, nor the subject of reasonable debate. It is just wrong."
The Illness Defense
Threading through every phase of the case is what Judge Moorer calls a "mysterious illness." Whenever the court raised concerns about timeliness or preparation — the missed jury demand, the discovery failures, the inability to question witnesses efficiently, the twelve-day trial — Eaton attributed the problem to this illness. In the summer of 2025, he claimed the same condition that had plagued him since the case began in 2022 prevented him from being fully ready to try it.
The court's response to this is direct and unsentimental. Moorer writes that he "concludes at this point that Mr. Eaton demonstrates an unwillingness or inability to meet the minimum standard of competence to practice law" and adds that "the Court finds that Mr. Eaton is either lying about his health situation or, if true, he is no longer capable of practicing trial law and should file for disability." This is not language judges use casually. It is a binary verdict: either Eaton has been deceiving the court for years, or he has a genuine medical condition that renders him unfit to practice — and in either case, he should not be practicing.
The opinion does not attempt to adjudicate which of those two possibilities is true. That question is now the Alabama State Bar's to answer. But the judge makes clear he has thought about it and finds neither scenario a defense.
The Math of Accountability
The $55,597 sanction was calculated from 218.9 hours of defense attorney time spent responding specifically to the misstatements of law — the fabricated citations, the false quotations, the erroneous authorities. Moorer notes that his own chambers spent a comparable number of hours on the same problem: cite-checking every reference, drafting orders, reviewing Eaton's responses to the show cause proceedings. The judge is effectively saying that the waste imposed on the court mirrors the waste imposed on opposing counsel, which is part of why he finds the requested fee amount reasonable.
The hourly rates — ranging from $110 for paralegal work to $359 for senior partner time — are, the court observes, actually below the rates it has recently approved as reasonable in other matters. This is not a punitive multiplier. It is straight compensation for time actually spent on a problem one attorney created for everyone else in the room.
Beyond the money, the reputational architecture of the sanction is notable. The requirement that Eaton attach this opinion to every active and future filing for a year is not a fine. It is a disclosure regime. Every future client, opposing counsel, and presiding judge will receive, as a matter of procedural record, a federal court's conclusion that their attorney was found to have submitted fabricated case citations and was recommended for a finding of incompetence. The opinion, once filed in the Federal Supplement, is permanently accessible to anyone who searches his name in a legal database.
What This Means Beyond One Case
The AI-hallucination sanctions wave has been building for three years. The most prominent early case involved two New York attorneys who submitted ChatGPT-generated briefs with fabricated citations in a personal injury case and were sanctioned by Judge P. Kevin Castel in 2023. Since then, similar cases have appeared in federal courts across the country, and bar associations have issued guidance warnings that attorneys remain personally responsible for the accuracy of every citation they file, regardless of how it was generated.
What makes the Moorer opinion different is not the AI angle but the comprehensiveness of the failure it documents. Most hallucination-sanction cases involve an isolated incident — an attorney who used an AI tool carelessly once and got caught. The Eaton case involves a pattern of professional failure across years of litigation, with the fabricated citations functioning as one data point in a much larger record of inadequate representation. The court is not primarily punishing the use of AI. It is punishing the sustained abdication of professional responsibility, of which the AI-generated citations are the most recent and most visible symptom.
Critics might note that courts themselves have been slow to establish clear, uniform standards for AI use in legal filings. Some jurisdictions require disclosure when AI tools are used; many do not. In the absence of consistent rules, attorneys may underestimate their verification obligations, particularly if they are already stretched thin or — as the Eaton case may illustrate — struggling with other pressures. The absence of a coherent national framework for AI in legal practice does not excuse what Eaton did, but it does suggest the profession has not yet built the guardrails to prevent the next version of it.
Critics might also observe that the harshest penalties in this case — the bar referral, the incompetence recommendation — are driven by the full pattern of conduct rather than the AI fabrications alone. An attorney who submitted a single AI-generated brief with erroneous citations, caught it immediately, corrected it, and cooperated fully with the court would almost certainly face a far lighter outcome. The severity here is a product of years of failures compounded, not of a single technological misstep.
Still, the piece underscores a structural point: the legal system's adversarial design provides a check that many other professional contexts lack. Opposing counsel caught the fabricated citations before the court did. The motion to strike forced the issue into the record. The show cause orders gave Eaton opportunities to respond that he declined to use. The system worked — slowly, expensively, and only after significant waste — but it worked. Whether that process is adequate deterrence for the broader profession remains an open question.
Bottom Line
The Eaton sanctions opinion is a detailed portrait of what happens when professional negligence is allowed to compound across years of litigation and then collides with a judge willing to document all of it in permanent, public, citable form. The AI-hallucination angle brought the case to wider attention, but the opinion's real subject is simpler and older: what courts owe to their own integrity, and what attorneys owe to the clients and institutions they serve. Judge Moorer's answer, rendered in 12,000 words and $55,597, is that both debts are real and both come due.