AIPolicyTracker
Menu

AI incident ·

Claude Console Reportedly Generated Phantom Legal Quotations in Trump Layoffs Court Filing

2 news reports Snapshot 7 Sep 2026

In brief

An AI system built by Anthropic and deployed by Binnall Law Group and Jason Greaves allegedly harmed Epistemic Integrity, Judicial Integrity and 3 others.

Risk domain
Misinformation False or misleading information
Occurred
Coverage
2 reportsMay 2026

What happened

A Binnall Law Group attorney reportedly used Anthropic's Claude Console to draft a motion to quash a subpoena in litigation over Trump administration layoffs, then filed a version containing "phantom" quotations that did not appear in the cited cases. The attorney apologized to U.S. District Judge Susan Illston, and the firm said it would add safeguards and training after calling the AI-generated errors unacceptable.

Laws that address this harm

Policy angle: Classified under Misinformation (False or misleading information) in the MIT AI Risk Repository taxonomy; 5 recorded instruments address this use case.

Matched from the record's risk domain and country to the instruments recorded here. A reviewer can correct the match in the repository (data/external/incident_overrides.yaml).

News reports (2)

Titles link to the original publisher; report text is not reproduced here.

  1. Notice of Declarations from Third-Party Joseph Guy's Counsel – Case Nos. 3:25-cv-3698-SI and 3:26-mc-80146-SI
    fingfx.thomsonreuters.com · Lindsay R. McKasson, Jason Greaves, Jesse R. Binnall
  2. Lawyer apologizes for 'phantom' AI quotes in Trump layoffs case
    reuters.com · Mike Scarcella

Who was involved

Alleged developer
Anthropic
Alleged harmed party
Epistemic Integrity, Judicial Integrity, Joseph Guy, Parties In Trump Administration Layoffs, Parties In American Federation Of Government Employees Afl Cio V. Trump

Classification (MIT AI Risk Repository taxonomy)

Risk domain
Misinformation
Causal entity
AI
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)

  • 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)

  • 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)

  • 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)

  • 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)

  • Untruthful Content

    "The LLM-generated content could contain inaccurate information"

    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)

Incidents in the same risk subdomain

All incidents in this subdomain

Source record: incident #1499 on the AI Incident Database · all 2 reports