Managing AI Risk in the Solicitors’ Profession

A profession changing in real time

According to Thomson Reuters, active use of generative AI among legal firms in the US, UK and Canada rose from 14% in 2024 to 26% in 2025[1]. Adoption is accelerating, but remains uneven, making the risk harder for insurers to assess.

For Paul Crilly, head of Recruitment and Solicitors at Pen Underwriting, AI now occupies the position cyber once did: a fast-moving issue that can improve how firms work while changing how insurers evaluate exposure.

AI governance becomes the differentiator

As AI spreads across legal work, governance is the clearest way for firms to show underwriters they’re managing risk. The question isn’t whether a firm uses AI, but how it controls that use.

“Misusing the wrong platform has the potential to be career-ending for solicitors. Firms need a documented AI policy naming approved tools, as well as staff training and senior leaders who own oversight and verify compliance,” says Crilly.

When firms cannot show basic controls, that can be regarded as a warning sign. It may point to weak supervision, no formal policy, unrestricted public tools, no audit trail and poor confidentiality controls. If someone enters sensitive client information into an AI platform, the firm can face data protection, regulatory and reputational problems before anyone realises.

When faulty AI output becomes a claim

In 2025, the High Court of England and Wales considered cases involving AI-generated or AI-assisted legal research. According to The Guardian’s reporting on the judgment, an 89 million-pound damages claim against a specific bank included 45 case-law citations, 18 of which were fictitious[2].

The court made it clear: lawyers must verify AI-generated research before putting it before a court. If they submit fake authorities, they may face professional consequences; if the error harms a client’s case, the firm may also face a negligence claim.

Phishing and business email compromise also remain persistent concerns. The Solicitors Regulation Authority, which regulates solicitors and law firms in England and Wales, continues to treat cybercrime as a priority risk. It has also warned law firms about AI-enabled deepfakes, including manipulated audio and video, being used to bypass identity checks, following an alert from the National Economic Crime Centre.

 

Considerations for brokers

AI is efficient, but this alone will not satisfy underwriters. The strongest firms will be able to show that AI supports human work rather than replacing professional judgement.

Paul Crilly, head of Recruitment and Solicitors at Pen Underwriting

 

While the market is still navigating AI-related claims, brokers can help solicitor clients understand how AI changes professional indemnity risk. Ahead of renewal, brokers can support clients by turning challenges into practical conversations: what AI tools are permitted, who is responsible for oversight, how client data is protected and how AI-generated work is reviewed before it reaches clients.

 

The opinions and views expressed in this article are those of the author and are provided for general guidance only. The author disclaims any liability arising from reliance on these opinions and encourages readers to consult more than one source before making decisions based on the information provided.