Financetoday
financialTelegram

AI Data Centers Could Strain Insurance Capacity, Aon's CEO Warns

Insurance·November 5, 2026

Aon chief executive Greg Case has warned that the rapid buildout of artificial intelligence data centers could run into a basic constraint: how much insurance the market is willing and able to write against them.

The core problem is concentration. A single large AI campus can represent billions of dollars of physical assets, specialized equipment and business interruption exposure, all in one location. When that much value sits in a small number of sites, insurers have less room to spread the risk across their books. Carriers manage this by capping how much they will cover in any one place, and once those caps are reached, the capacity left for new projects shrinks.

Case's point reaches beyond insurance itself. Lenders and equity investors generally want a project to be insurable at a workable cost before they commit capital. If coverage becomes scarce or expensive, developers may face higher premiums, retain more risk on their own balance sheets, or accept tighter terms. Any of those shifts can raise the cost of building a facility, or push some developments back until the risk is better understood.

The concern fits a wider pattern in which investment in fast-moving technology outpaces the risk models and reinsurance capacity meant to support it. Data centers at this scale are still relatively new, and they are filled with costly hardware that can be hard to replace quickly, which makes loss estimates difficult to pin down.

For developers and financiers, the practical takeaway is that insurance is moving earlier into project planning. Site selection, the spread of exposure across regions and the structure of risk transfer may all matter more as the sector grows. Whether that leads to slower projects, higher financing costs or new insurance products will depend on how the market adjusts.

Reporting based on an external source.