The rapid construction of increasingly large and complex data centres has triggered a surge in demand for insurance, fuelled by growing investment in artificial intelligence (AI) infrastructure, energy systems and semiconductor manufacturing.
According to Swiss Re’s Sigma 2026 report, AI hyperscaler capital expenditure is expected to reach $750 billion in 2026, while AI-related investment accounted for nearly 40% of US GDP growth during the first three quarters of 2025.
The global investment boom in AI infrastructure is reshaping commercial risk pools. Re/insuring data centres is complex across construction – involving physical perils, subcontractor dependencies, and delays and operational phases.
New AI data centres can cost more than $20 billion to build before equipment is installed, with lenders increasingly requiring very large insurance limits before financing projects.
Swiss Re explained that modern AI data centres are larger and more complex to engineer than traditional builds. Constructed as campuses with dense, interdependent systems, they concentrate risk within single sites.
Once GPUs, tenants, and services are established, value and operational complexity surge, making continuous availability paramount and elevating business interruption, loss of rent, and service disruptions to critical risks, analysts noted.
These capital-intensive projects require advanced cooling, high-voltage power, backup infrastructure, sophisticated hardware, and robust security software.
Insurance cover for these large projects is driven by their multi-billion financing, creating demand for very high insurance limits at a single data centre location.
Financing institutions require limits that cover the full cost of construction, even as maximum probable loss scenarios are much lower.
The re/insurance industry can currently only support a fraction of this limit at competitive rates for traditional construction risk policies, the report highlighted.
“This unprecedented wave of investments into the data centres requires insurance, most of the time, across the entire life cycle of a project. We estimate that the demand could be, from an insurance premium standpoint, of up to $90 billion on an aggregated basis until 2030. These are premiums. If you think about the capital, the capacity associated with this demand, we’re talking about 200, 300, 400 billion of capital that is required to support these projects,” Ivan Gonzalez, Chief Executive Officer of Swiss Re Corporate Solutions stated.
Data centres are also increasingly being built in high natural hazard locations. According to the report, around 40% of US data-centre capacity could be located in significant-to-very-high tornado-risk zones, while more than a quarter could face elevated large-hail exposure.
This vulnerability comes at a time when insured losses from natural catastrophes are rising by 5‒7% annually on average in real terms over the long term.
Commenting on the report, Gonzalez highlighted that in order to build resilience against the new risks , a holistic view across prevention, insurance, and risk financing is needed.
To navigate these challenges, Swiss Re’s risk engineering team developed the 3W model, Gonzales explained, which focuses on watts, water, and waiting time.
When it comes down to watts Swiss Re tries to estimate the energy required, how it is generated, and its long-term sustainability. Evaluating the availability of local water supplies for the cooling of many of these data centres is also vital.
Additionally, addressing delays in both construction and repairs is important. While the physical shells of data centres are relatively straightforward to build, their components are highly globalised, making supply chain risk a critical vulnerability.
Gonzales concluded: “Three things are key, risk engineering is becoming increasingly important to have the ability to assess how these are being constructed. The second one is alternative risk transfer, so the traditional insurance solutions are not always enough when it comes down to this type of very large risks.
“And when we think about parametric solutions, for example cat bonds, many of the corporations behind these projects have captives, so captive management is also quite important. And the last one is credit and surety, because you need to think about the financing, the construction, the operational risk that is associated with it.”
This boom extends beyond AI, as the world is entering a broader capital investment cycle across data centres, energy infrastructure, and advanced manufacturing.
This shift, Swiss Re stated, is creating new pools of insurable risk, and expanding demand across property, engineering, cyber, liability and business interruption insurance.
