Digital health companies are integrating AI faster than the insurance market has worked out how to cover it. Beazley's 2026 Digital Health & Wellness report puts that gap at the centre of its findings. The survey of 600 executives across Europe, North America, and Asia carries direct implications for how this class of business is placed and advised. Beazley writes digital health insurance and has a direct commercial interest in the findings it publishes - the underlying data is worth reading with that context in mind, though the claims analysis draws on its own book over a decade rather than a single survey cycle.
The most striking data point is not the AI adoption rate. It is what is actually driving claims. Beazley's healthcare claims book over the past decade ranks medical negligence as the most frequent and most severe cause of loss, with improper supervision second. Yet the risks dominating executive attention are cyberattacks and workforce competency.
The report identifies what it calls a liability chain: a single AI-assisted patient harm event can generate claims across medical professional liability, cyber, technology errors and omissions, and general liability simultaneously.
Marc Martin, product leader for digital healthcare at Beazley, said the liability exposure in AI-enabled care extends well beyond the treating clinician. "In the event of patient injury, claims are likely to target all parties involved, including tech companies, physicians, and clients, due to the interconnected nature of AI-enabled healthcare solutions," he said.
That convergence is already shifting how digital health firms buy insurance. The proportion purchasing a single, tailored multi-risk policy has risen from 40% in 2024 to 53% in 2026, according to the report. Separate and siloed placements leave gaps when a claim spans multiple lines. As AI embeds deeper into clinical workflows, those gaps are harder to defend.
Who bears liability when AI-assisted care goes wrong has also reached the highest levels of the legal system. The UK Jurisdiction Taskforce, chaired by the Master of the Rolls, published its Legal Statement on Liability for AI Harms in July. Its central conclusion is that English law is capable of addressing AI-related harm without new legislation, starting from a simple premise - AI has no legal personality and cannot itself be liable. Responsibility falls on the developers, deployers and users in the chain. That allocation sharpens the question of where coverage actually applies.
Workforce competency concerns are well-founded. Thirty-six per cent of executives cite professional misrepresentation risk, and Beazley's claims data confirms improper supervision as the second most frequent cause of loss.
Cyber risk, however, tells a different story. Ranked highly by 32% of executives, it generates claims whose frequency currently outpaces their severity. Breach of contract, intellectual property claims and AI-related miscommunication receive comparatively little executive attention despite featuring prominently in Beazley's claims analysis.
That mismatch between what executives are focused on and what is actually generating losses is the report's most useful finding for a broker. A digital health client whose risk management investment is concentrated on cyberattack prevention may have underweighted the professional liability and contractual exposure that sits closer to the top of the claims frequency table.
Against that backdrop, Beazley's survey records a significant shift in how digital health firms choose their insurers. Fast and reliable claims handling has overtaken price and coverage as the leading purchasing factor. The report points to a claims environment that is more adversarial, harder to investigate and more expensive to resolve. Privacy litigation, AI-related errors, and social media misinformation are converging into increasingly complex multi-party disputes. That shift in purchasing priorities changes the weight of the claims conversation in placement.
The change coincides with a broader market movement. Insurers are actively revising policy wording to address silent AI exposure - policies that neither expressly cover nor exclude AI-related risks. For digital health clients integrating AI across diagnosis, triage, treatment and administrative functions, silent wording is a live placement issue. It spans professional liability, cyber and technology lines simultaneously, and the Beazley data suggests those clients are already beginning to demand clarity on it.
For brokers advising digital health clients, the report points to three specific conversations worth having before the next renewal. First, whether the current programme structure responds across all lines when a single AI-assisted incident generates claims in parallel. Second, whether the executive team's risk priorities match what is actually driving frequency in the claims book. Third, whether existing policy wording takes a position on AI-related loss - and if not, what that silence means for coverage when a claim arrives.
Beazley's 2026 survey was conducted by Opinion Matters between March 9 and 17. Respondents represented health and wellness practitioners, software and platform providers, health technology companies, mHealth providers, and telehealth and telemedicine businesses.
