Across 21 jurisdictions, US lawyers face varied AI guidance; courts have sanctioned attorneys over filings containing AI-hallucinated citations. A 2026 survey found 69% of legal professionals use general-purpose AI at work, while just 9% of firms enforce a written policy.
In the federal civil-rights case Withers v. City of Aberdeen, a Mississippi judge earlier this year sanctioned four lawyers after both sides submitted filings with AI-generated hallucinated case citations. Two attorneys were barred from appearing in the district for two years and fined; two local attorneys were also sanctioned. In Wadsworth v. Walmart, a Wyoming federal judge in February 2025 sanctioned three attorneys after a firm AI platform cited eight nonexistent cases. The drafter was fined $3,000 and lost admission to practice in the case; two others were each fined $1,000.
The American Bar Association’s Formal Opinion 512 says lawyers using generative AI remain subject to duties of competence, confidentiality, supervision and candor to a tribunal. Ethics lawyer Cliff McKinney said verification failures underlie most AI problems; New York City Bar ethics chair Michael E. Salzman also cited risks involving fabricated citations, incorrect legal analysis, false factual claims and confidentiality or attorney-client privilege. Lawyer John Rajaee said his firm uses AI to summarize medical records, build treatment timelines, organize documents and translate complex medical or insurance language, with a human checking anything that matters.
