Most corporate AI deployments are uninsured. That was the opening charge Philippe Cotelle brought to more than 3,000 risk and insurance professionals at the FERMA Forum 2026 in Rotterdam.
Cotelle is president of the Federation of European Risk Management Associations (FERMA) and head of cyber insurance management at Airbus Defence & Space. In his opening address, he named two structural problems the industry has been too slow to confront. The first is a coverage gap that leaves AI-driven business operations largely exposed. The second is an annual renewal cycle that bears little relation to how businesses actually plan.
The more pressing concern centred on what Cotelle called the AI paradox. Artificial intelligence is becoming embedded in the operational fabric of organisations, yet the term AI is, in his words, "almost non-existent" in policy wordings. Businesses on their digital journey face uncertainty about whether their insurance responds to AI-related losses at all.
The market is already moving to close that gap, though not always in the direction policyholders would prefer. In the US, in January 2026, the Insurance Services Office (ISO) introduced a generative AI exclusion in commercial general liability policies. Some major carriers have since added explicit AI exclusions across liability, directors and officers, and errors and omissions lines. Clients renewing policies that looked familiar at first glance may find AI-related losses now sit outside coverage, a shift that has often gone unremarked in renewal discussions.
Generative AI-related lawsuits in the US surged by 978% in the five years to 2025, according to a report by Gallagher Re produced with MIT.
"This lack of clarity around AI in insurance contracts is unsustainable for corporates," Cotelle said. "Insurance has a critical role to play in supporting this transition and we need to secure the foundations of that role now."
He proposed that FERMA lead a multi-stakeholder dialogue on AI risk. Industry-wide coordination of that kind has historically preceded major shifts in how new exposures get underwritten. It would also require brokers to engage clients on coverage terms many will not yet have reviewed.
The second argument was less urgent in timing but more structural in scope. The insurance cycle is annual. Corporate operational planning runs on three-to-five-year horizons. A factory cannot be built or a product developed for a single year. The annual renewal forces a short-term rhythm onto a relationship that businesses need to treat as long-term.
"A partnership that only lasts when everything is going well is not a partnership, it is merely a transaction," Cotelle said.
The frustration is not new among European risk managers, and brokers have heard it from clients seeking more stable, predictable coverage arrangements. What Cotelle added was a sharper name for the consequence. The current model, where either party can exit when conditions shift, creates an inherent fragility in the relationship. He called for a long-term mutual commitment capable of absorbing volatility rather than terminating the arrangement because of it.
Multi-year policies have existed in certain lines for some time, but their uptake remains limited. Brokers have historically derived commission income from annual renewal activity, and that has been cited as one structural barrier to wider adoption of longer-term arrangements.
The Rotterdam forum drew attendees under the theme Widening the Lens: From Risk Insight to Strategic Foresight. FERMA represents 24 national risk management associations across 23 European countries. Cotelle signalled that the AI dialogue and the push for longer-term insurance structures would be priorities for his tenure.
He also referenced Open Sesame, a FERMA partnership with the World Business Council for Sustainable Development (WBCSD). The project aims to develop a unified financing framework for climate-related prevention and adaptation, mobilising capital before losses occur rather than after.
