Legal Malpractice Claim Frequency Rose in 2025 for First Time in Years as Insurers Cite Uncontrolled AI Use as a Growing Cause of Professional Liability
Insurers providing professional liability cover to law firms reported an increase in the frequency of legal malpractice claims in 2025 — the first such rise in several years — with uncontrolled use of artificial intelligence cited as a contributing factor, according to this year's legal professional liability insurance survey by EPIC Law Firm Group. The finding is commercially significant because it links the accelerating adoption of AI tools in legal practice directly to a measurable uptick in the claims that erode law firms' professional indemnity insurance records and, over time, their premiums. Legal malpractice — negligence in the provision of legal services — has historically been driven by missed deadlines, conflicts of interest, and inadequate advice. The emergence of AI-related causes marks a structural shift in the risk profile of legal professional liability. The survey's reference to 'uncontrolled' AI use is precise: it is not AI per se that generates liability, but AI deployed without adequate supervision, quality-checking, and governance frameworks. For firms that have rolled out AI drafting and research tools without corresponding verification protocols, the insurance data now provides a concrete, quantified signal that the risk is real and is already materialising in claims. This development sits alongside the Pinsent Masons SRA self-referral reported elsewhere in today's briefing, together painting a consistent picture of the professional conduct and liability exposure accumulating across the profession.
Why this matters
The link between uncontrolled AI use and rising legal malpractice claim frequency is the kind of hard data point that converts a theoretical governance concern into a concrete client advisory issue. Law firms advising other professional services businesses on their AI governance frameworks can now point to insurance market data as evidence that inadequate AI oversight creates quantifiable liability risk. For law firms themselves, the data will sharpen the internal business case for investment in AI governance infrastructure — supervision workflows, output verification protocols, and training — before insurers begin repricing premiums or imposing coverage exclusions for AI-related incidents.
On the Ground
A trainee supporting a law firm's internal AI governance programme, or advising a client on their legal AI deployment, would be drafting AI governance policy documents, reviewing technology licence terms for AI vendor tools to identify liability allocation provisions, and preparing regulatory impact assessment memos mapping the firm's AI tool deployment against the SRA's published guidance on technology and competence.
Interview prep
Soundbite
Insurance data showing AI driving malpractice claims means governance frameworks are no longer optional — they're an underwriting condition waiting to happen.
Question you might get
“How should a law firm structure its AI governance framework to manage professional indemnity risk, and what specific oversight mechanisms would reduce the likelihood of an AI-related malpractice claim?”
Full answer
A legal professional liability insurance survey by EPIC Law Firm Group found that malpractice claim frequency rose in 2025 for the first time in years, with insurers specifically citing uncontrolled AI use as a contributing cause. This matters because it transforms AI governance from a reputational concern into a measurable insurance and liability risk, giving managing partners a concrete business case for investing in supervision protocols and output verification. The wider picture is a profession where AI tool adoption has outpaced governance frameworks — a gap that is now visible in claims data as well as in conduct incidents like the Pinsent Masons SRA self-referral. Firms that act now to implement structured AI oversight should be better positioned when insurers begin differentiating premiums based on firms' AI governance maturity.
My notes
savedLoading…