Research
This research examines how a wave of AI hallucinations in court filings is reshaping professional liability underwriting. We show how lawyers have emerged as the first target for sweeping AI exclusions, why legal AI vendors are pushing liability squarely onto the firms deploying their tools, and how the same wording trend is already spreading to architects and medical facilities. It delivers fast insight on where silent AI coverage is starting to break, how carriers like Hamilton and Berkley are recalibrating E&O wordings, and what's at stake for professionals as exclusions ripple across the wider market.
Subscribe to get comprehensive access to our exclusive analysis in the full platform.
Law firms could be the first hit with tightened AI wording and exclusions on professional liability policies as AI-related incidents proliferate across the industry, sources told Insurance Insider US.
Last month, global firm Sullivan & Cromwell came under scrutiny after one of its lawyers submitted a court filing with more than 40 hallucinated AI errors and false citations.
It is just one incident in a growing trend of AI-driven legal mishaps. AI researcher Damien Charlotin has tracked over 370 court decisions since June 2023 of cases where lawyers were accused of presenting hallucinated AI content. These decisions represent only a small portion of AI-related incidents in total.
Similarly, AI liability MGA Testudo has tracked 720 US incidents so far where an attorney was sanctioned or disciplined for filing AI hallucinated content. The highest recorded fine was $110,000.
“It’s inevitable that it will be a problem everywhere, but for whatever reason, lawyers just can’t seem to keep it in check,” one broker said.
Legal AI vendors are not being named in negligence cases in instances where a law firm has relied on a GenAI output because vendors transfer the liability to the deployer of their technology, according to Testudo. In other words, the law firm absorbs liability by reviewing outputs before they reach third parties.
Exclusions aren’t widespread at this point, but brokers canvassed by this publication reported seeing some sweeping AI exclusions on lawyers’ errors and omissions (E&O) policies with the expectation that there’s more to come.
Number of legal decisions in cases where generative AI produced hallucinated content
Delving E&O policies cover legal fees, settlements and damages when professional service providers are sued for things like negligence, errors or inadequate services. They are commonly purchased by law firms, accountants, architects, financial advisors and other professionals in regulated industries.
Precursor to broader trend
This emerging trend in the legal space could be a predecessor to more widespread AI exclusions in the professional lines market, with some carriers already implementing them across multiple industry verticals.
Hamilton is one of the insurers leading the charge on broad AI exclusions in its professional liability forms, introducing exclusions as far back as 2024.
Berkley has also formulated a broad exclusion for incidents “based upon, arising out of, or attributable to” the actual or alleged use, deployment, or development of artificial intelligence.
Besides law firms, brokers also reported seeing broad AI liability exclusions on architects’ and medical facilities’ policies, with one describing the trend as “concerning.”
Despite widespread AI use in financial services, one underwriter said that they have not yet seen exclusions emerge in FI, as insurers are currently at the stage of probing clients on whether that AI use is purely for internal purposes versus how much of it bleeds into output to customers.
Instead, professions likely to produce client-facing materials face more immediate risk from AI hallucinations, they said.
Brokers speaking to this publication said most E&O policies across all industry segments are still silent on AI coverage, though some insurers are starting to implement sub-limits as one form of affirmative AI coverage.
The consensus is that while silent coverage may be acceptable now, it is not likely to last.
“It will just take one large case where policy is silent, and the insurer actually declines it for people to ask for affirmative coverage,” one source said. “Even if an organization isn’t deploying AI at a company level, what’s stopping an associate from using ChatGPT or Claude on their own?”
Longstanding relationships between insurers and their corporate clients have prevented underwriters from putting sweeping AI exclusions in place at this point, according to a source at a large retail broker.
“Carriers are very mindful of not tarnishing relationships by just throwing a broad exclusion on a policy without looking and understanding exactly what they're looking to exclude versus cover,” they said. “There's certainly a lot of talk about it, and underwriters are absolutely trying to get a better handle on the risk they have within their portfolio.”
By Olivia Royle
May 13, 2026
Request your free trial today to unlock our complete intelligence platform.