Who's Liable When AI Gets the Law Wrong?

Judge Sharion Aycock was clearly unhappy. On June 8, the senior federal judge sanctioned lawyers on both sides of a civil case, fining them and referring them for bar discipline. And Aycock, who sits on the U.S. District Court for the Northern District of Mississippi, took the further—and even more extraordinary—step of disqualifying the lawyers from the case and canceling the trial.

The issue? Both sides used artificial intelligence tools in preparing their briefs, and AI delivered fake case citations that counsel had failed to check. "This case presents the court with an unusual scenario—attorneys for both litigants engaged in similar sanctionable conduct," Aycock wrote. "Their practice of blindly relying on technology resulted in the hallucinatory citations contained in their respective filings."

Aycock's order is one of the most recent examples of growing anxiety and anger among judges around AI-related snafus in court. And judges are not alone. Clients appear to be losing patience as well. One recent survey of major insurers found that for the first time, legal malpractice claims over AI-related mistakes are on the rise.

If a new lane for malpractice suits is opening, many law firms may find themselves woefully unprepared. Professional liability insurance policies written just a few years ago may not cover artificial intelligence-related errors, potentially leaving the firm exposed if its AI usage goes awry.

A Bad Citation

Even legal players that presumably have the institutional resources to catch potential AI errors are running into trouble. On July 16, the U.S. Department of Justice ran afoul of a judge for court filings that appeared to use generative AI and included nonexistent case law. Reuters reported that a chief judge in the U.S. District Court for the Western District of Michigan had "raised concerns" about AI usage after she was unable to find a government-cited case in a court filing in an immigration suit.

"Although the court will not presently impose sanctions for this conduct, it goes without saying that the government must ensure its future filings with this court do not include nonexistent case law," Judge Hala Jarbou wrote.

According to Jarbou, the government said a case, Taylor v. Hott, was from the U.S. Court of Appeals for the Sixth Circuit and justified its arguments concerning a bond decision for a man detained by U.S. Immigration and Customs Enforcement. While researching the case, the judge could not find it on the page the government cited in the Federal Appendix. That page contained a commercial arbitration opinion, not one on immigration bond issues, the court said.

"In its research, the court was unable to identify a Sixth Circuit case with the caption Taylor v. Hott, or any federal case containing the quoted language. Thus, it seems this citation was likely produced by generative artificial intelligence," Jarbou said.

An Escalating Problem

Incidents like those in Aycock's and Jarbou's courts are not isolated ones. According to a global database tracking AI hallucination cases, U.S. lawyers, judges, experts, and paralegals have filed court documents with AI hallucinations in 495 cases since April 2023. Just under half of those cases, 222, have occurred in 2026 alone. The database, which has been cited by the federal courts, is maintained by Damien Charlotin, a French lawyer and data scientist.

And sanctions against lawyers have been increasing. Some 96 lawyers have been sanctioned for AI misuse since the beginning of 2025—compared to two in 2023 and 2024 combined, the database shows. Sanctions in 2026 are outpacing those issued last year: as of mid-July, 39 lawyers had been sanctioned for AI hallucinations, compared to 13 during the same period in 2025.

Firms large and small have had hallucination issues:

• In July, the U.S. Court of Appeals for the 11th Circuit said it would refer a Florida solo practitioner to its Committee on Lawyer Qualifications and Conduct for submitting a series of briefs that include fake, AI-generated case citations. By outsourcing his legal work in a commercial aviation employment case to an AI algorithm, the court said the lawyer "violated his ethical duties to both his clients and this court." Reuters reported that the lawyer apologized to the appeals court last year, saying the fake citations stemmed from research oversights.
• In April, Wall Street titan Sullivan & Cromwell was forced to apologize to a federal judge for submitting a brief to a U.S. Bankruptcy Court in New York that included several AI-created errors, including fake case citations. According to The New York Times, the firm gave the court a ledger detailing errors that spanned three pages and totaled around three dozen — many involving citations to seemingly imagined passages from real cases.
• Last year, three lawyers from Butler Snow were reprimanded and referred for bar discipline by an Alabama federal judge over court filings with AI-fabricated case citations. The firm had been hired to defend Alabama and other jurisdictions in prison litigation. "Fabricating legal authority is serious misconduct that demands a serious sanction," the judge said, removing the lawyers from the case and ordering them to share the sanctions order with clients, opposing lawyers, and judges in all of their other cases.

Rising AI Malpractice Claims

From a liability perspective, AI errors are helping fuel a new category of malpractice claims. In May, the law firm arm of EPIC, a national insurance brokerage and consulting firm, reported in its 16th annual Lawyers' Professional Liability Claims Survey that seven of 13 companies offering malpractice insurance to law firms reported an increase in AI-related claims over the past year. Those 13 companies provide 80% of the malpractice coverage to Am Law 200 law firms.

"What is most concerning is that AI-related malpractice exposure has moved from theoretical to real," Eileen Garczynski, a principal of EPIC Law Firm Group and author of the survey, said in a news release. "The duty of competence cannot be delegated to technology."

In an interview with Insurance Business, Garczynski added that underwriters have been consistently concerned about AI and were asking law firms questions. "But they weren't actually seeing claims," she said. "This year, when I asked whether they were seeing AI-related claims, the answer was 'yes.' "

Insurers told EPIC that malpractice claims were rising overall, and the cost of defending claims was increasing as well. AI "stands apart," Insurance Business noted, because it is a new avenue for claimants—one likely to push underwriters toward closer scrutiny of firms' governance, training and controls. Insurers, Garczynski said, want evidence that firms understand the risks and are taking meaningful steps to address them.

Silent No More: The Coverage Gap Explained

Even as AI-related risk grows, law firms may face a potential snag if they wish to rely on their current malpractice policies to cover potential claims. If a policy was written prior to 2023, it is unlikely to include specific language about artificial intelligence — meaning that if a claim occurs, a firm may or may not be covered.

Specifically, firms may be relying on what's known as "silent" AI coverage, where AI-related issues are implicitly covered under traditional policies without being explicitly named. "This 'silent' coverage is similar to how early cyber risks were handled under standard policies before dedicated cyber insurance existed," WTW, a risk management company and insurer, wrote in a December 2025 report. An AI incident may be covered by current policies—cyber, liability, and so on—even if they don't specifically mention AI.

But the insurance landscape is changing quickly. In a June article, Fenwick noted that a growing number of insurers have begun to narrow coverage for AI-related risks. Policyholders may need to pay close attention at renewal, since AI-related exclusions may not show up in one obvious place — narrower language can appear in revised base forms, AI-specific endorsements, new exclusions, definitions, application questions, underwriting file positions, or restrictive carve-backs.

Insurers may decline to underwrite AI-related risks altogether, move them outside a policy's base coverage, raise premiums, or subject policyholders to greater underwriting scrutiny and carve-outs around AI outputs, decision-making, system behavior, or third-party tool use, Fenwick said.

The Road Ahead

In most cases, law firms carry professional liability, cyber liability, and errors and omissions insurance—and firm leaders may assume one of those would respond to an AI-related claim. But artificial intelligence may fall into coverage gaps in each. As one law firm AI consultant recently put it, when AI hallucinates a citation, a malpractice policy should respond, but the insurer might call it a technology failure; when AI leaks client data, a cyber policy should respond, but the insurer might instead call it a professional services issue.

For firm leaders, that means several concrete steps are worth taking now:

• Review current policies to understand exactly what is—and isn't—covered, and prepare for more rigorous underwriter questions about AI usage going forward. Wisconsin Lawyers Mutual Insurance recently advised policyholders that attorneys using AI should be ready to explain what tools they use, how those tools are supervised, and what safeguards are in place to reduce the chance of a malpractice claim.
• Consider specialized AI-related insurance products as they emerge. Lloyd's of London, for instance, recently introduced new coverage for losses caused by malfunctioning AI tools.
• Watch compliance risk alongside insurance risk. Hundreds of bills targeting AI practices have been introduced in U.S. state legislatures, and international regulators are active too—the EU's new AI Act requires businesses to assess the risks posed by their AI systems, and common legal AI tools for document review, research, or contract analysis could be classified as high-risk under that framework given their influence on legal outcomes.

Law firm leaders, in other words, should be on guard for both unforeseen compliance and insurance risk simultaneously.

"Many companies are facing increasing commercial pressure to adopt AI to remain competitive without adequately considering the extent this adoption changes their risk profile," one large firm partner recently told Bloomberg,

Frequently Asked Questions

• Does malpractice insurance cover AI-related errors? It depends on the policy. Many policies written before 2023 contain no AI-specific language, which means AI incidents may fall under "silent" coverage rather than being explicitly addressed. Insurers are increasingly adding AI-specific exclusions or endorsements at renewal, so firms should confirm their actual coverage directly with their carrier rather than assume it either way.
• Are lawyers actually being sanctioned for AI mistakes in court? Yes, and at a rising rate. A global tracking database has logged AI hallucinations in 495 U.S. court filings since April 2023, with 222 of those in 2026 alone. Ninety-six lawyers have been sanctioned for AI misuse since the start of 2025—up from just two in the two years prior.
• Are malpractice claims related to AI actually increasing? Yes. EPIC's 2026 Lawyers' Professional Liability Claims Survey found that more than half of the 13 major insurers surveyed—which together provide 80% of malpractice coverage to Am Law 200 firms—reported a rise in AI-related claims over the past year.
• What can law firms do to close the AI insurance coverage gap? Firms should review existing malpractice, cyber, and E&O policies for AI-specific language or exclusions, prepare for more detailed underwriter questions about AI governance and safeguards, and consider newer AI-specific insurance products as that market develops.

--

David L. Brown is a legal affairs writer and consultant, who has served as head of editorial at ALM Media, editor-in-chief of The National Law Journal and Legal Times, and executive editor of The American Lawyer. He consults on thought leadership strategy and creates in-depth content for legal industry clients and works closely with Best Law Firms as senior content consultant.