Judge Reportedly Disqualifies Butler Snow Lawyers Following Purported Use of ChatGPT-Fabricated Citations in Alabama Prison Litigation

Three Butler Snow attorneys defending a former Alabama corrections commissioner were reportedly disqualified after two motions cited five non-existent cases generated via ChatGPT. Judge Anna Manasco is reported to have publicly reprimanded the lawyers, referred them to the Alabama State Bar, and ordered broad disclosure; she declined to sanction the firm. An external review of 2,400 citations across 330 filings reported no further errors.

Risk classification

• Primary risk domain: 3 Misinformation
• Primary risk subdomain: 3.1 False or misleading information

The AI system generated fabricated legal citations, which constitutes discrete false and misleading information that was subsequently submitted to a federal court.

Additional risk subdomains

• 5.1 Overreliance and unsafe use: The attorneys over-relied on ChatGPT's output, failing to verify the citations before submitting them to the court.

Causal factors

• Entity: AI
• Intent: Unintentional
• Timing: Post-deployment

The risk was initiated by ChatGPT generating fabricated citations (hallucinations), which was an unexpected outcome of the user's query.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Risk Level 3: Limited Risk. The system involved is ChatGPT, a general-purpose generative AI chatbot. According to the definitions, chatbots and AI-generated content pose a moderate risk and are classified as Limited Risk.

AI system and alleged parties

• AI system: ChatGPT (OpenAI)
• AI purpose: Writing Assistant; Question Answering
• Behaviour type: Assistant
• Alleged developer: OpenAI
• Alleged deployer: William Lunsford, William Cranford, Matthew B. Reeves, Butler Snow
• Alleged harmed parties: William Lunsford, William Cranford, State of Alabama, Matthew B. Reeves, Judicial integrity, General public of Alabama, Frankie Johnson, Epistemic integrity, Butler Snow, Alabama Department of Corrections

Harm severity

Severity ratings are based on this scale.

Highest direct severity in any category: Minor. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.

• Physical: direct Negligible, indirect Negligible
• Infrastructure: direct Negligible, indirect Negligible
• Property: direct Negligible, indirect Negligible
• Financial: direct Negligible, indirect Minor
• Environmental: direct Negligible, indirect Negligible
• Malicious content: direct Negligible, indirect Negligible
• Differential treatment: direct Negligible, indirect Negligible
• Civil rights: direct Negligible, indirect Negligible
• Democracy: direct Negligible, indirect Negligible
• Privacy: direct Negligible, indirect Negligible
• Psychological: direct Negligible, indirect Negligible
• Epistemic: direct Minor, indirect Minor
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Financial

Reported: Yes, the law firm Butler Snow incurred financial costs from hiring another firm to audit filings, though the exact dollar amount was not explicitly reported.

Directly caused: The report does not explicitly state a specific dollar amount of direct financial loss caused by the incident.

Indirectly caused: The law firm Butler Snow incurred indirect financial costs from hiring Morgan, Lewis & Bockius to conduct an independent review of over 2,400 citations across 330 filings, though the exact cost was not disclosed.

Inferred additional harm: It is likely the firm spent tens of thousands of dollars on the independent audit and lost potential revenue from the disqualification of the three attorneys.

Epistemic

Reported: Yes, the report explicitly describes the fabrication of five legal citations by ChatGPT, which were submitted in federal court filings.

Directly caused: ChatGPT generated five completely fabricated case citations that did not exist or did not stand for the propositions cited, which were then submitted to a federal court.

Indirectly caused: The submission of fake citations misled the court and opposing counsel, requiring independent verification by the judge and opposing attorneys, and disrupting the legal proceedings.

Inferred additional harm: The incident contributes to a broader erosion of trust in legal filings and the integrity of the judicial system as AI-generated hallucinations increasingly appear in court documents.

People affected

• Occurrences reported: 1
• People reportedly harmed: 3
• People reportedly exposed: 3

Potential causes

Management

• Lack of Pre-Filing Controls: Management lacked active technical controls to block unauthorized AI tools.

Technology

• AI Hallucinations: ChatGPT fabricated non-existent legal citations and case details.

Human Factors

• Failure to Verify: Reeves failed to verify ChatGPT citations via Westlaw or Pacer due to haste.
• Policy Non-Compliance: Reeves knew the AI policy limitations but chose not to comply with them.
• Blind Trust by Co-Signers: Cranford and Lunsford signed filings without checking the citations.

Process and Methods

• Inadequate Review Process: The firm's filing process allowed unverified AI citations to be submitted.

Regulatory Environment

• Insufficient Deterrents: Prior modest court sanctions failed to deter systemic misuse of generative AI.

Information quality

• Classification confidence: High
• Reason for confidence: The reports provide clear, consistent, and detailed accounts of the incident, including the specific law firm, the lawyers involved, the judge's rulings, the nature of the AI-generated fabrications, and the resulting sanctions. There is no conflicting information regarding the core facts of the AI's role or the outcomes.

Attorneys at a US law firm submitted court filings containing fabricated citations generated by ChatGPT. The incident is a matter of professional misconduct and judicial administration with negligible national security implications.

• Overall national security impact: Negligible
• Scope: Single nation
• Primary target: United States
• Alleged perpetrator: Individual

Threat characteristics

• Imminence: Long-term. The incident represents a long-term professional standards and verification challenge rather than an active national security crisis.
• Autonomy: Human-controlled. The AI acted as a writing assistant, and humans retained full control over whether to submit the generated text to the court.
• Novelty: Established threat. Generative AI hallucinations in legal filings are an established phenomenon with multiple documented prior instances.

Impact by dimension

• Physical security: Negligible. The incident involved legal document fabrication and has no relation to physical security, kinetic capabilities, or critical infrastructure.
• Information security: Negligible. There was no compromise of intelligence, classified data, or state-sponsored information warfare; it was an isolated case of professional negligence.
• Sovereignty: Negligible. While occurring within a federal court, the incident is a routine legal ethics violation managed through standard judicial and bar procedures, posing no threat to state authority.
• Economic security: Negligible. The financial impact was limited to a private law firm hiring auditors, with no broader implications for national economic or technological security.
• Societal stability: Negligible. The incident did not threaten social cohesion, civil liberties, or population safety, representing only an isolated professional misconduct issue.