Attorney Sepideh Ardestani was sanctioned $1,500 over AI-hallucinated citations in a filing

Plaintiff's attorney Sepideh Ardestani filed a motion for reconsideration in Tercero v. Sacramento Logistics containing two nonexistent case citations, ten fabricated quotations, and twelve misattributed legal propositions. When confronted, Ardestani denied using AI and provided inconsistent explanations that the court found not credible. U.S. District Judge Dena Coggins imposed a $1,500 sanction and directed the clerk to refer the matter to the State Bar of California.

The Work Justice Law Firm · Incident Sep 9, 2025 · Indexed Jun 4, 2026 · 2 sources

A language model fabricated plausible-sounding case citations that the attorney filed without verification and then lied to the court about the source.
What
Plaintiff's attorney Sepideh Ardestani filed a motion for reconsideration in Tercero v.
Incident date
Sep 9, 2025
Who
The Work Justice Law Firm
Failure mode
Hallucination
AI surface
Chatbot
Severity
Medium

What happened

On September 9, 2025, U.S. District Judge Dena Coggins sanctioned attorney Sepideh Ardestani $1,500 after finding she filed a motion for reconsideration containing two entirely fabricated cases (Klein v. Avante Group, Inc. and Gersh v. Anglin), ten fictitious quotations attributed to real cases, and twelve cases cited for propositions they did not support. The court noted the citations bore the hallmarks of AI hallucinations and that Ardestani repeatedly gave inconsistent and not credible explanations when given multiple opportunities to correct the record. Judge Coggins characterized the conduct as akin to contempt of court and directed the clerk to serve the order on the State Bar of California for disciplinary proceedings.

What broke inside the model

Failure path · mode profile · Hallucination
  1. 01 · TriggerA user asks for a fact, a citation, or a figure.
  2. 02 · Model stepThe model writes a fluent, confident answer.
  3. 03 · Control gapNothing ties the claim back to a real source.
  4. 04 · FailureA fabricated fact ships as if it were verified.
  5. 05 · ConsequenceThe false claim reaches a customer, a court, or the public.

Confidence holds, and even spikes, as the claim detaches from any source.

A generative AI chatbot likely produced fabricated judicial opinions that appeared plausible but had no basis in actual case law, a well-known hallucination pattern where language models synthesize convincing but fictitious legal citations. The attorney failed to verify any of the AI-generated citations against primary sources before filing them with the court, and then compounded the failure by providing inconsistent and misleading explanations about her research methods when the errors were exposed.

Public visibilityMedium
Regulatory exposureActive
Customer impactFew customers
Financial impactDisclosed
Time to disclosureWeeks
  1. Court FilingOrder Sanctioning Plaintiff's Counsel, Sepideh Ardestani, Tercero v. Sacramento Logistics, No. 2:24-cv-00953-DC-JDPcourthousenews.com
  2. PressAttorney Is Sanctioned for Suspected AI Fabricated Citationsnews.bloomberglaw.com
Permalinkhttps://failureindex.ai/failures/attorney-sepideh-ardestani-sanctioned-1-500
CitationAI Failure Index. "Attorney Sepideh Ardestani was sanctioned $1,500 over AI-hallucinated citations in a filing" (FI-0128). Realm Labs. https://failureindex.ai/failures/attorney-sepideh-ardestani-sanctioned-1-500 (indexed Jun 4, 2026).
Share cardA branded image of this record for posts and slides.

Data fields CC-BY 4.0, prose citation permitted. Incident ID FI-0128. Full dataset at /data.

Note from Realm Labs, the Index steward

How Realm would have caught this

Controls for this failure mode
  • Prism
  • OmniGuard
  • AI Detection & Response (AIDR)

A runtime layer that watches the model's internal state can flag the moment a model commits to a claim it has no support for, and hold or reroute the response before it reaches a user. Realm reads those signals in real time rather than grading the transcript after the fact.