This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
In a breach of contract case in Mississippi, four lawyers were sanctioned and removed from the proceedings after submitting court filings that included hallucinated legal citations generated by AI tools. The presiding judge noted that the attorneys failed to verify the accuracy of the AI-generated content, violating civil procedure rules. The incident highlights the ongoing challenges regarding the use of generative AI in legal research and documentation.
In Withers v. City of Aberdeen, a federal judge in Mississippi reportedly sanctioned four lawyers after filings from both sides contained purportedly hallucinated legal citations linked to unverified AI use. The court said Kathleen Wilson used First Drafts to draft a filing and Kathryn Williams used an AI legal-research tool; all four attorneys were removed from the case and fined.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
AI systems that inadvertently generate or spread incorrect or deceptive information, which can lead to inaccurate beliefs in users and undermine their autonomy. Humans that make decisions based on false beliefs can experience physical, emotional or material harms
Also classified as
The lawyers blindly relied on the AI-generated citations without verifying their authenticity, leading to professional sanctions.
The AI systems generated fabricated legal citations (hallucinations) which were submitted as factual evidence in court filings, misleading the court.
AI system
Due to a decision or action made by an AI system
Unintentional
Due to an unexpected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed
The incident was caused by the AI system generating hallucinated citations, which was an unexpected and unintentional outcome of its deployment.
Added categories peak at 2: Minor (Epistemic Harm, Direct)
Four lawyers in Mississippi were sanctioned for submitting AI-generated fake legal citations in a civil lawsuit. The incident represents a professional misconduct and technology integration issue with negligible national security implications.
The AI systems functioned as assistants designed to help draft legal documents and conduct research, but required independent verification by the lawyers, as noted by the developer First Drafts.
Risk Level 3: Limited Risk. The AI systems involved are drafting and research assistants that generate text, falling under transparency obligations for AI-generated content.
The AI system 'First Drafts' is described as an AI-powered program for drafting legal documents, which serves as a writing assistant for legal professionals.
The report provides clear, detailed, and uncontradicted facts about the lawyers, the specific AI tools used, the sanctions imposed by the judge, and the nature of the AI hallucinations.
None
Technology
AI Hallucinations
AI-powered program generated nonexistent legal citations.
Limited Data Scope
In-house tool lacked Mississippi law but was used for research.
Data Inputs
Incomplete Training Data
AI models lacked access to verified, up-to-date regional case law.
Human Factors
Ignorance of AI Risks
Attorney did not know that generative AI could produce fake citations.
Blind Reliance on AI
Attorneys trusted AI outputs without performing manual verification.
Process and Methods
No Verification Step
Attorneys failed to check the authenticity of citations before filing.
Management
Absence of AI Policy
Solo practice lacked guidelines or policies for AI tool usage.
Disregard of Policy
Firm partner ignored existing policy requiring verification of research.
Regulatory Environment
Rule 11 Violation
Attorneys certified false information as factual in court filings.
AI system involved
Identified by automated extraction; can differ from the alleged developers above.
