A Probate Appellant Trusted AI Citations He Never Checked. Two of Them Did Not Exist.

What happened

Jeffrey Kotchka filed an opening brief in the Arizona Court of Appeals as a self-represented appellant challenging the administration of his late mother's estate. The brief contained eight legal citations. When the court reviewed them, six were found deficient. Two of those six cited cases that did not exist at all.

The dispute centered on the appointment of Kotchka's sister, Kim Dineen, as personal representative of the estate. The court found that Kotchka had made no effort to verify the citations before filing. That finding mattered both procedurally and practically: fabricated case law does not support legal arguments, but it does require opposing counsel to identify and document each false citation, and that work costs money.

The court affirmed Dineen's appointment as personal representative and ordered Kotchka to pay the attorney fees she incurred specifically as a result of addressing the false citations. The sanction was calibrated to the harm: not a general fee award, but compensation tied directly to the wasted effort the bad citations created. For a self-represented litigant without legal training, the episode follows a recognizable arc. AI drafting tools generate plausible-sounding citations, complete with case names, court designations, and reporter formats that look authoritative on the page. Without independent verification against a legal database, there is no way to know whether a cited case exists.

The Arizona case sits inside a growing body of similar incidents. Courts across jurisdictions have found fabricated citations in briefs filed by both self-represented parties and practicing attorneys who used AI drafting tools without checking their outputs. The pattern is consistent: the tool hallucinates, the filer trusts, and the court eventually discovers the gap. The sanctions that follow land on the filer, which is appropriate, but they do not illuminate what the tool returned, what confidence it expressed, or whether any warning accompanied the output.

That last part is the accountability gap the case makes visible. Courts reviewing these incidents have no mechanism to establish what tool produced the citations, what the tool returned alongside them, or whether a verification step was available and skipped or simply unavailable by design. Without a provable record of what a system generated and what it conveyed to the user at the moment of generation, the full chain of failure stays invisible. What remains is the document, the sanction, and a finding of no effort to verify, which is accurate as far as it goes but stops well short of what the record could, and should, contain.

Reported impact

Affected parties

Not publicly disclosed

Harm type

Not publicly disclosed

Scale

Not publicly disclosed

Financial impact

Not publicly disclosed

Regulatory action

Not publicly disclosed

Classification

Organization

Not publicly disclosed

AI system

Not publicly disclosed

Industry

Not publicly disclosed

Country

Not publicly disclosed

Provider

Not publicly disclosed

Incident type

Not publicly disclosed

Relevant governance controls

Governance control mapping is not available for this record.

• No controls mappedNot publicly disclosed

Control mapping is analytical. It does not state that any control would have prevented the incident.

Sources and evidence

This record was researched and written by the Index. The event is also catalogued in the following database, which is listed for cross-reference.