An Attorney Filed a Federal Motion Full of Case Quotes That Did Not Exist
What happened
A Binnall Law Group attorney used an AI drafting tool to prepare a motion to quash a subpoena in federal litigation over Trump administration layoffs, then filed a version of that motion containing "phantom" quotations, passages attributed to real cases that did not actually appear in those cases. The filing landed before U.S. District Judge Susan Illston, who received a document that looked legally sourced and was not.
The motion was part of a case challenging the administration's federal workforce reductions, a high-profile dispute drawing substantial public and judicial attention. The attorney used an AI console to produce the draft, carried the output forward into the filed version, and the fabricated quotations made it through without being caught before submission. Whether the attorney read the citations closely or assumed the tool had produced accurate text is not established in the record.
When the problem surfaced, the attorney apologized directly to Judge Illston. Binnall Law Group acknowledged that the errors were unacceptable and stated the firm would implement safeguards and additional training to prevent the same failure from recurring. The apology and the corrective pledge represented the firm's public response; the full scope of the court's reaction is not specified in the record.
This was not an isolated failure of a novel technology but a variation on a problem courts have been documenting for several years. Lawyers have submitted AI-generated briefs containing citations to cases that do not exist, quotes from opinions that were never written, and legal standards that were invented whole. The professional obligations around legal research, including the duty to verify every citation before filing, predate AI tools by decades. Those obligations did not change when drafting moved to a text-generation interface.
The accountability gap the incident exposes is not primarily technical. A text-generation tool that produces plausible-sounding but false quotations is a known failure mode, and the legal profession has been warned about it repeatedly. What is missing is verification infrastructure: a provable record of what a system produced, what a human reviewed, and what was independently confirmed before a court filing was signed and submitted. Without that record, the gap between a draft and a verified document is invisible, and courts are left discovering the difference only after it matters.
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
Relevant governance controls
Governance control mapping is not available for this record.
- No controls mapped
Not 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.