AI Risk Wire

Cases › made things up › LR-2026-0477

Judge fines and sanctions two city lawyers over AI-generated citations

Preliminary Why: Publisher not yet named · subject to change

Status: action reported, no fix

Severity minor
Reported 1 Apr 2026
Developer and model ChatGPT

No source states: loss or records, agent tool, could do, deployer, industry.

Severity minor: Test: contained to one deployer (the city Law Department) with no third party touched; the sources state fines and resignations, not any wider effect.

A law department in the City of New Orleans used ChatGPT to help research a motion in a federal civil rights lawsuit. The filing included nine made-up case citations generated by AI, and the attorney said he did not check whether the cases were real before adding them to the brief. A supervising deputy city attorney reviewed the motion before filing but did not question the strange formatting of the fake citations. The federal judge found the citations during the case and ordered the attorneys to appear in court. At a March 19 hearing one attorney admitted the AI use. The court ruled that the filing violated Rule 11, imposed fines of $250 and $1,000, and issued a formal warning to a third official. Both attorneys resigned, and the office adopted a written AI policy effective March 27, 2026.

What happened

1 · What went wrongThe attorney did not check whether the AI-generated cases were real and did not read them before filing.
2 · What it reached or didThe filed motion included nine made-up case citations generated by AI.
3 · How it was caughtThe federal judge discovered the false citations and ordered the attorneys to appear in court.

What was done

The court fined Harris $250 and Roquemore $1,000, and both attorneys later resigned.

Risks

Read from the sources; not yet reviewed by a person.

Controls

Not yet read for controls. Controls that mitigate the primary risk:

Application

Grounding and fact checks · Human approval for high-impact actions · Output filtering and redaction · People told they are dealing with AI

Similar incidents

All cases of this kind ›

Evidence

No source states: what the AI was doing, when it happened, when it was found, how many records were exposed, the direct loss, a count of the other parties affected, any court case, what set it off.

Show the full account: 3 items from the primary sources

What happened, in order

  1. 2026-01 Motion containing the false citations was filed in the Gentry lawsuit.
  2. 2026-03-19 At a hearing, Harris admitted he used ChatGPT to help research the motion.
  3. 2026-03-27 The City Attorney's Law Department AI policy took effect.
Show every field, with its source and grade

The AI system

Model
ChatGPT
C one secondary source states it

What happened

Earliest source we hold
2026-04-01
D inferred by us from the described behaviour, reason shown
Cause
a human check that existed was skipped or failed
C one secondary source states it
What went wrong
made things up
C one secondary source states it
primary risk
RK14
D inferred by us from the described behaviour, reason shown

Impact

Severity
minor
D inferred by us from the described behaviour, reason shown
Why this severity
Test: contained to one deployer (the city Law Department) with no third party touched; the sources state fines and resignations, not any wider effect.
D inferred by us from the described behaviour, reason shown

Context

Country
United States
C one secondary source states it

Notes on gaps

The exact date of the resignations is not stated. The supervisor's failed review and the missing check by the attorney are not clearly mapped to a listed control; I7 is used for the missing verification and the supervisory review is not coded. The summary describes the third official's warning as formal, as the source states, and characterises the supervisor's review neutrally.

LR-2026-0477 · version 1 · published 2026-11-01 after automatic checks; not yet checked field by field by a person · JSON