The rapid rise of AI in law firms is exposing a mismatch between tool capabilities and lawyers’ professional duties. With inconsistent guidance across jurisdictions and limited firm policies, AI-generated “hallucinations” have already led to court sanctions in high-profile cases. Experts stress that ABA Opinion 512 leaves lawyers responsible for accuracy, confidentiality and supervision, and firms must pair AI with strict human review.
Patchwork AI Rules Leave Lawyers Vulnerable As Courts Punish ‘Hallucinated’ Filings
A stock image shows a judge striking a gavel in a courtroom.
The rapid adoption of artificial intelligence in law firms is exposing a growing gap between what these tools can produce and the professional responsibilities lawyers must uphold. As AI-generated "hallucinations" produce fabricated citations and other factual errors, attorneys and firms must navigate a fragmented set of rules while remaining accountable for what they file in court.
Fragmented Guidance Across Jurisdictions
A paper published in August by the New York City Bar Association reviewed guidance on AI across 21 jurisdictions and found no single, industrywide standard. Instead, courts and professional organizations are issuing their own rules and recommendations, leaving practitioners to manage inconsistent obligations depending on where they practice.
That inconsistency matters as use grows: a 2026 survey of more than 1,300 legal professionals found that 69% were using general-purpose AI at work, yet just 9% of firms had a written, enforced AI policy.
Courts Begin To Punish Hallucinations
The risks are no longer theoretical. In Withers v. City of Aberdeen, a federal civil-rights case in Mississippi, a judge sanctioned four lawyers after both sides filed briefs that included case citations generated by AI that did not exist. Two attorneys were barred from appearing in the district for two years and fined; two local lawyers also received sanctions.
In February 2025, a federal judge in Wyoming sanctioned three attorneys in Wadsworth v. Walmart after a filing produced with the firm’s internal AI platform cited eight nonexistent cases. The lawyer who drafted the filing was fined $3,000 and lost his admission to practice in the case; two other attorneys were fined $1,000 each.
Existing Professional Duties Apply
These rulings underscore duties already embedded in professional standards. The American Bar Association’s Formal Opinion 512 states that lawyers who use generative AI remain bound by existing obligations — including competence, confidentiality, supervision and candor toward tribunals.
Cliff McKinney of Quattlebaum, Grooms & Tull PLLC, who has written about AI and ABA Opinion 512, told Newsweek that inconsistent guidance creates compliance challenges, particularly around confidentiality and attorney–client privilege.
“Most problems with the use of AI boil down to the failure of the attorney to verify the accuracy of the AI’s results,” McKinney said.
Michael E. Salzman, chair of the New York City Bar’s Professional Ethics Committee, warned of similar risks: fabricated citations, flawed legal analysis and false factual assertions — plus confidentiality exposures when lawyers rely on general-purpose AI systems.
How Firms Are Responding
Many firms are attracted to AI for its ability to speed routine work: summarizing records, organizing documents, drafting timelines and translating technical language. But leading practitioners emphasize human oversight for any work that matters to a case or client.
John Rajaee of RMD Law described his firm’s approach: they use AI for tasks such as summarizing medical records and organizing documents, but apply a strict verification rule for anything submitted externally.
“AI is a very good paralegal, but it's not a lawyer,” Rajaee said. “At our firm, we have a simple rule: If it matters, a human checks it.”
Where The Industry Goes From Here
The central tension is between assistance and accountability. AI can save lawyers hours of routine work, but it cannot assume the professional obligations attached to that work. Firms that wish to capture AI’s efficiencies must pair technology with clear policies, robust human review, and training that enforces verification, confidentiality safeguards and supervisory responsibilities.
For now, regulators and courts are signaling that mistakes produced by AI will be treated as lawyer errors. The challenge for the legal industry is not whether to use AI, but how to use it safely — preserving human judgment, verification and accountability while benefiting from technological efficiency.
Help us improve.
