AI Hasn't Triggered Flood of Legal Malpractice Claims, but Insurers Watching
AI hasn’t triggered flood of legal malpractice claims, but insurers watching
In an August 25, 2026 article written for The Maryland Daily Record, Berkley Select's SVP of Professional Liability Al Roberts contributes and emphasizes that while AI can be a valuable tool, attorneys remain accountable for the accuracy of their work and the protection of client information.
Lisa Burden//August 25, 2026//
When lawyers began filing court papers containing fictitious cases generated by artificial intelligence, judges weren’t the only ones paying attention. So were the companies that insure lawyers.
Although some legal malpractice insurers say they have not yet seen a significant wave of claims directly attributable to generative AI, interviews with professionals suggest the industry is quietly preparing for a future in which AI becomes a routine part of legal practice — and potentially a new source of professional liability.
For now, insurers say they’re taking a measured approach. They aren’t broadly raising premiums or excluding AI-related claims from coverage. Instead, they are urging law firms to establish policies governing AI use, protect confidential client information and ensure lawyers independently verify AI-generated work before relying on it.
“The most important story is not just efficiency,” said Chris Newbold, chief operating officer and executive vice president of ALPS, a Montana-based lawyers’ professional liability insurer that sells in Maryland. “It moves to responsibility.”
AI can help attorneys work faster and more efficiently, Newbold said, but it also can increase “the speed and the scale” at which mistakes occur if lawyers fail to exercise independent professional judgment.
“The lawyer still is going to own the judgment, the verification and the work product at the end of the day,” he said. “The underlying risk here is not new. Lawyers have always had the duty to verify their work.”
Early warning signs — not a wave of claims
The legal profession has seen a growing number of sanctions imposed on lawyers who filed AI-generated briefs containing fabricated citations or quotations. In May, for example, a judge scolded lawyers at a Severna Park firm.
Yet insurance professionals interviewed by The Daily Record, all of whom sell policies in Maryland, say those highly publicized incidents have not translated into a significant increase in malpractice claims.
“We’re not seeing a lot of what I would call direct AI claims at this point,” Newbold said. Instead, AI is appearing as “part of the fact pattern” in matters involving inaccurate citations or insufficient verification.
He described those incidents as “an early signal rather than a clear trend.”
RELATED: AI creates gray area for MD law schools, students
Lee Norcross, president of L Squared Insurance Agency, reached a similar conclusion. L Squared, a nationwide, Michigan-headquartered insurance broker, specializes in lawyers’ professional liability insurance.
Although his agency has written extensively about AI hallucinations, “we as an agency have not seen any claims directly related to AI use,” he said.
Al Roberts, who heads professional liability for Chicago’s Berkley Select, declined to discuss his company’s claims experience but agreed that lawyers who rely solely on AI-generated research without verifying it risk both professional discipline and malpractice exposure.
Lawyers, not AI, remain responsible
For lawyers, the issue extends beyond embarrassment or adverse publicity.
Lawyers’ professional liability insurance generally pays to defend covered malpractice claims and, when appropriate, indemnify lawyers for covered losses. However, malpractice policies typically don’t cover court-imposed sanctions, disciplinary penalties or fines.
That distinction has become increasingly important as judges sanction attorneys who submit briefs containing fictitious authorities generated by AI.
“Attorneys using AI need to remember that they are the ones that can be sanctioned and possibly disbarred for not verifying an AI-generated brief,” Norcross said. “AI does not have a law license to lose.”
Roberts made much the same point. Lawyers can’t defend themselves by claiming they did not know AI systems hallucinate or by arguing that the AI assured them its research was accurate.
“You can’t get away with that,” Roberts said.
Attorneys are licensed professionals, he said, and remain responsible for the accuracy of every filing submitted under their names.
Governance of AI, not a prohibition on AI
Perhaps the strongest theme emerging from the interviews was not concern about AI itself but concern about how law firms manage it. None of the insurance professionals suggested that lawyers should avoid AI use altogether. Instead, they repeatedly emphasized governance – the policies, procedures and oversight that determine how AI is used inside a law firm.
“I think we’re in this interesting transition period where adoption might be moving faster than governance,” Newbold said.
Overreliance on AI, rather than AI itself, appears to be a strong concern. AI should function much like a junior associate or paralegal — it can assist lawyers, but it can’t replace their judgment.
“We need to guide the profession to focus on governance, diligence and accountability because at the core, from our perspective, AI can be a very valuable assistant,” Newbold said.
Roberts agreed that firms should establish formal AI use policies rather than allowing lawyers and staff determine their own AI practices. However, he cautioned, the existence of a policy alone won’t protect a firm after a mistake occurs. Instead, a policy is intended to prevent those mistakes in the first place.
Confidentiality and closed systems
Accuracy is not the only concern. Lawyers also must understand what happens to confidential information entered into an AI system. Confidentiality, Roberts observed, is an important facet of legal representation.
He said firms should consider using closed or controlled AI environments designed to prevent client data from being disclosed, retained for training or exposed outside the firm.
Lawyers also must understand the ethics requirements of each jurisdiction in which they practice, he said. States have taken different approaches to questions such as when lawyers must disclose AI use and what safeguards are required. Roberts noted that he is licensed in Illinois and in that state, lawyers don’t have to tell a client they are using AI.
That inconsistency presents challenges for national law firms and for insurers attempting to evaluate firms operating in multiple jurisdictions.
“How do you underwrite it if one state requires you to do something and another state doesn’t?” Roberts asked. He described that as a question insurers will have to confront as state standards continue to develop.
Newbold similarly said firms should avoid placing confidential information into free, publicly available AI platforms and should look for products with strong security protocols and controlled environments.
That issue may be especially difficult for solo and small-firm lawyers who lack the technology budgets of the largest firms. Newbold nevertheless said increasingly affordable legal technology may give smaller firms meaningful opportunities to use AI safely and compete more effectively.
Premiums, exclusions and underwriting largely unchanged
Despite widespread speculation about how insurers will respond to AI risk, the professionals interviewed described a market that is watching and gathering information rather than making sweeping changes.
Norcross said AI has not directly affected malpractice premiums. Instead, he said, insurers themselves increasingly are using AI to verify information submitted on insurance applications and potentially to assist in claims investigations.
Norcross said he is aware of one nonstandard surplus-lines insurer with an endorsement excluding AI use but described that as “the exception, not the rule.”
Roberts noted that Berkley Select does not have any exclusionary language in place and is not currently asking law firms supplemental underwriting questions about their AI use, though both practices could change as the technology and claims experience evolve.
Newbold said ALPS is likely to move toward asking more detailed questions about how firms use AI. Because insurer applications and forms are subject to state regulatory approval, however, such changes can’t always be made immediately.
He added that insurers will likely examine both ends of the process: how firms describe their AI use when applying for coverage and what claims data eventually reveal about actual losses.
The use of AI itself might not determine whether a firm represents a heightened risk, Newbold said. The more important questions will involve how the firm governs that use and whether it has controls designed to prevent mistakes.
Looking ahead
The consensus among insurance professionals is that AI will become a permanent part of legal practice. The question is not whether lawyers will use it, but whether they will use it responsibly.
For now, insurers say they are less concerned with the technology itself than with how lawyers supervise it, verify its output and protect client confidences.
As Newbold put it: “The legal question here is not whether AI is going to be used. It’s whether it’s going to be used with discipline.”
Al Roberts is the Senior Vice President of Lawyers Practice Liability and Affinity Programs at Berkley Select.
Link to article: https://thedailyrecord.com/2026/08/25/ai-hallucinations-legal-malpractice-insurance/
Originally published in The Maryland Daily Record - August 25, 2026
Disclaimer:
Berkley Select is a member company of W. R. Berkley Corporation, a Fortune 500 Company. The views expressed here are those of the author and do not necessarily represent the views of Berkley Select or W. R. Berkley Corporation.