Skip to main content
Citation Safe
← All teardowns

Compiled by Citation Safe (Digital Empire Holdings LLC) from the court’s published sanction order (decided September 16, 2026). Quotes and case details come from the order itself — check the primary source before relying on them.

Sanction teardown · Fla. 4th DCA, USA · 2026-09-16

Lisandrillo v. Palozzi

What happened

In Fla. 4th DCA, USA, a filing relied on an unnamed/unconfirmed AI tool to help draft legal argument. The court identified the following problems with the citations in that filing:

  • Misrepresented (Legal Norm)
    The filing asserted that a motion to disqualify automatically stayed all trial-court proceedings. The court characterized this as a recurring AI-hallucinated argument and held that the assertion was legally mistaken.
  • Misrepresented (Exhibits & Submissions)
    The motion to disqualify alleged that the judge systematically muzzled counsel and refused to allow a proffer. The court reviewed the hearing transcript and found those allegations contrary to the record, making them insufficient for disqualification.
  • Misrepresented (Exhibits & Submissions)
    The filings alleged that the court continued proceedings while counsel’s audio malfunctioned and repeatedly misidentified counsel as “Mr. Soroka.” The official transcript did not support those claims; it showed that the court alerted counsel when she was muted and contained no such misidentifications.
  • Misrepresented (Exhibits & Submissions)
    The motion asserted that the court admonished counsel for failing to cross-examine witnesses, using the phrase “The bind that creates, participate and get shut down on the one hand . . . decline and get admonished on the other.” The court found the passage to be legal gibberish and noted that the record contained no citation or quotation supporting the alleged admonishment.
  • Misrepresented (Exhibits & Submissions)
    The petition and motion repeatedly claimed that counsel was prevented from presenting a defense, making arguments, preserving issues, or moving to disqualify during the contempt hearing. The court found these claims contradicted by the transcript, which showed that counsel had opportunities to act but did not do so.
  • Misrepresented (Legal Norm)
    The filings asserted that the trial court could not hear from the guardian ad litem because no written report had been filed. The court rejected the argument as baseless, explaining that Florida law permits a guardian ad litem to address the court and make oral recommendations, particularly where the report could not be completed because contact with the child had been cut off.
  • Misrepresented (Case Law)
    The motion relied on an unpublished decision in an unrelated case to accuse the trial judge of departing from the essential requirements of law. The court held that the decision had no precedential value and was irrelevant to the proceeding.

Which AI tool

an unnamed/unconfirmed AI tool. Note: Charlotin's public database records tool attribution only where a court order, brief, or reporting on the matter states it explicitly; "unidentified" or "implied" means the record indicates AI use but does not name a specific product — we do not guess.

Outcome

Petition denied; counsel ordered to show cause, without using AI, why sanctions including possible Florida Bar referral should not be imposed.

Additional detail

The court found that counsel’s filings appeared to contain AI-generated components, including convoluted and frivolous arguments, alleged events contrary to the official record, and legal gibberish. The court expressly noted that the cited cases were real and that the matter did not involve hallucinated law or false legal citations. It rejected the petition and disqualification motion, emphasizing that a judge cannot be disqualified based on AI-hallucinated events or allegations contradicted by the record. Counsel was ordered to show cause within ten days, without using AI, why sanctions or referral to the Florida Bar should not follow.

How Citation Safe would have caught this

Citation Safe runs three deterministic layers before a brief is filed: (1) does the citation exist against CourtListener's database of published opinions, (2) if quoted, does that exact language appear in the source, (3) does the cited case actually support the proposition it is cited for. Fabricated case citations fail Layer 1. Fabricated or misattributed quotations fail Layer 2 even when the underlying case is real. Misrepresented holdings — a real case cited for a proposition it does not support — are the target of Layer 3. None of these checks involve asking another language model whether the citation looks right; they are lookups and text-matches against the actual source, which is why a hallucinated citation has to survive a direct lookup against the authoritative source — not another model's opinion — to earn a VERIFIED stamp; our measured false-verify rate is published live at /quality.

Check a brief before you file it → · See our live false-verify rate

Source: https://www.damiencharlotin.com/documents/3069/Opinion_2026-2262.pdf, via Damien Charlotin's public AI Hallucination Cases Database (CC0).