AIPolicyTracker
Menu

AI incident #1453 ·

Attorney in Fletcher v. Experian Information Solutions, Inc. Reportedly Submitted Reply Brief with Purportedly AI-Generated Material Misrepresentations

Open on the AI Incident Database 4 news reports Snapshot 7 Sep 2026

What happened

In Fletcher v. Experian Information Solutions, Inc., the Fifth Circuit found that attorney Heather Hersh had used generative AI to draft a substantial portion of a reply brief that contained fabricated legal support and factual representations. The court first issued a show-cause order and later imposed a $2,500 sanction, linking the filing errors to unverified AI-generated output.

Only the incident metadata is stored here. The underlying news reports are on the AI Incident Database (CC BY-SA 4.0); use the links above to read them.

News reports (4)

Coverage catalogued by the AI Incident Database. Titles link to the original publisher; the text is not reproduced here.

  1. US appeals court orders lawyer to pay $2,500 over AI hallucinations in brief
    reuters.com · Nate Raymond · AIID #7141
  2. Fletcher v. Experian Info Solutions: Appellate Sanctions for AI-Hallucinated Briefing and Lack of Candor Under FRAP 46(c) and Inherent Authority
    casemine.com · CaseMine · AIID #7449
  3. Fifth Circuit Sanctions Opinion Gives Practical Advice for AI Use
    texaslawbook.net · David Coale · AIID #7448
  4. Fifth Circuit Sanctions Lawyer for AI-Generated Brief
    lalegalethics.org · Dane S. Ciolino · AIID #7447

Who was involved

Alleged deployer
Heather Hersh
Alleged harmed party
United States Court Of Appeals For The Fifth Circuit, Judicial Integrity, James Fletcher, Experian Information Solutions Inc., Epistemic Integrity

Classification (MIT AI Risk Repository taxonomy)

Risk domain
Misinformation
Causal entity
Human
Intent
Unintentional
Timing
Post-deployment
Harm level
—
Sectors
—
Countries
—

Risk entries describing this failure mode

Entries from the MIT AI Risk Repository coded to subdomain 3.1.

  • Pursuing Consistent Context

    "LLMs have been demonstrated to pursue consistent context [129]–[132], which may lead to erroneous generation when the prefixes contain false information. Typical examples include sycophancy [129], [130], false demonstra...

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  • Knowledge Gaps

    "Since the training corpora of LLMs can not contain all possible world knowledge [114]–[119], and it is challenging for LLMs to grasp the long-tail knowledge within their training data [120], [121], LLMs inherently posse...

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  • Hallucinations

    "LLMs generate nonsensical, untruthful, and factual incorrect content"

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  • Faithfulness Errors

    "The LLM-generated content could contain inaccurate information" which is is not true to the source material or input used

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  • Defective Decoding Process

    In general, LLMs employ the Transformer architecture [32] and generate content in an autoregressive manner, where the prediction of the next token is conditioned on the previously generated token sequence. Such a scheme...

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  • Untruthful Content

    "The LLM-generated content could contain inaccurate information"

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  • Factuality Errors

    "The LLM-generated content could contain inaccurate information" which is factually incorrect

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  • Noisy Training Data

    "Another important source of hallucinations is the noise in training data, which introduces errors in the knowledge stored in model parameters [111]–[113]. Generally, the training data inherently harbors misinformation....

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

Incidents in the same risk subdomain

All incidents in this subdomain