Gallagher Re has launched a digital risk practice that consolidates AI liability, cyber, data centre, and digital risk engineering into a single advisory unit. The move responds to a growing market problem: technology failures now generate losses that cut across multiple insurance lines at once, leaving reinsurers with accumulation exposures that don't fit neatly into existing product categories.
Ian Newman leads the practice as global head of digital risk, alongside his existing role as global head of cyber. Freddie Scarratt, previously global deputy head of insurtech, takes on the AI liability lead position. Luca Drane has been appointed data centres lead. The three appointments pull expertise that was previously spread across separate functions within Gallagher Re's global products and practices platform into one team.
The firm said the practice will help clients translate technology exposures into underwriting, accumulation, and capital insights across property, casualty, and specialty classes.
The accumulation problem
The structure reflects a recognised gap in how the industry measures technology-driven risk. A single failure can hit cyber policies, property portfolios, business interruption cover, and casualty placements at the same time, but those exposures are often tracked separately.
The July 2024 CrowdStrike outage illustrated the scale of that gap. A faulty software update affected 8.5 million Windows devices globally. Guy Carpenter estimated insured losses at between US$300 million and US$1 billion. CyberCube Analytics put the figure between US$400 million and US$1.5 billion. Parametrix calculated total damages to Fortune 500 companies at approximately US$5.4 billion.
The difference between economic loss and insured loss in that event is exactly the kind of coverage shortfall the new practice is designed to address.
The data centre exposure
A Swiss Re Institute sigma report published in March projected global insurance premiums tied to data centres rising to US$24.2 billion by 2030, up from US$10.6 billion today. Capital spending by the five largest cloud providers is forecast to exceed US$600 billion in 2026, with roughly 75% of that spend directly tied to physical AI infrastructure housed in large data centres. Construction costs for a single facility can exceed US$20 billion before technology installation.
The Swiss Re Institute noted that large data centres concentrate multiple tenants and insured interests behind shared critical systems. Power, cooling, and fire protection failures can generate multiple concurrent claims from a single event. Separate insurance programmes for different site components can obscure total exposure within a reinsurer's portfolio.
Scarratt co-authored a Gallagher Re report on AI model risk in March that identified AI model failures as a source of aggregation risk across ceded portfolios, spanning existing cyber, casualty, and errors and omissions coverage.
AI liability as a standalone discipline
The practice's AI liability lead position reflects demand from cedants for reinsurance guidance on a class without settled policy language or established loss history. Munich Re's fourth Global Cyber Risk and Insurance Survey, published in May, found 63% of C-level executives wanted to buy insurance against AI-related risks. The same survey put the global cyber insurance market at US$15 billion and found 71% of C-level respondents identified AI as the technology most relevant to their business, up from 62% in 2024.
"Clients and markets are increasingly seeking guidance on the implications of AI and digital dependency, not only within cyber but across a wide range of business lines," Newman said.
Why this matters for insurance professionals
For brokers and underwriters, the practical takeaway is that technology risk can no longer be treated as a cyber-only conversation. The CrowdStrike event showed that a single software update can generate claims across property, casualty, and business interruption lines simultaneously. When structuring programmes for clients with significant digital dependencies or data centre exposure, the accumulation risk across those lines needs to be quantified as a portfolio issue, not line by line. Reinsurance buyers should expect their brokers to ask how AI model failures and technology concentration sit within their overall risk transfer strategy.
Your membership also unlocks:
