Overview
Swiss Re’s latest sigma report, published Wednesday, estimates that hyperscale operators will increase AI-related capital expenditure to $750 billion in 2026, up from about $500 billion in 2025. The report highlights how this surge in spending for data centres and related infrastructure will heighten demand for insurance and other risk-transfer solutions across multiple commercial lines.
Insurance industry response and market risks
Ivan Gonzalez, CEO of Swiss Re Corporate Solutions, framed the rapid deployment of data centres as a phenomenon that creates multi-layered and interconnected exposures. He said these require innovative responses from both insurers and capital markets to manage accumulation risk that spans financial markets.
"These interconnected exposures call for solutions that go beyond traditional insurance, combining risk engineering, alternative risk transfer and financing to help businesses invest with greater resilience," Gonzalez wrote in the report.
Economic and inflationary considerations
Jérôme Haegeli, Swiss Re’s group chief economist, characterised the pace and scale of AI-related infrastructure investment as without parallel in the post-war period. Haegeli warned that, although AI investment could drive positive growth effects, it may also add upward pressure to inflation. That inflationary impact, he said, could alter the claims environment and push up costs for insurers underwriting those risks.
Haegeli also set out the medium-term spending outlook contained in the report, saying AI-related investment is projected to reach roughly $1.6 trillion over the next five years. He described that flow of capital as a source of growth for the wider economy, for both physical and digital assets, and for insurance markets that will be called on to underwrite the associated risks.
Supply-chain and construction exposures
Swiss Re’s analysis notes that while the actual construction of large data centres is often straightforward, the underlying supply chain for a single U.S. facility can be highly complex. Ivan Gonzalez pointed out that materials and components for one data centre can originate from as many as 90 countries, a feature that increases supply-chain risk and potential accumulation of exposures.
Implications
The report implies a multifaceted adjustment will be required across the insurance value chain - from underwriting and risk engineering to alternative risk transfer and financing mechanisms - as hyperscale spending on AI infrastructure gathers pace. The combination of concentrated physical assets, global supply chains and sizeable capital flows creates accumulation and inflation-related pressures that market participants will need to address.
Key points
- Hyperscalers are projected to spend $750 billion on AI in 2026, up from an estimated $500 billion in 2025.
- Swiss Re says AI-related investment could total about $1.6 trillion over the next five years, supporting growth for the economy and insurance markets.
- The expansion of data centres raises complex accumulation and supply-chain risks that may require solutions beyond traditional insurance.
Risks and uncertainties
- Accumulation risk across financial markets tied to concentrated data-centre assets and large capital flows, affecting insurers and capital providers.
- Potential inflationary pressure from large-scale AI spending that could change the claims environment and increase costs for the insurance sector.
- Supply-chain vulnerability from sourcing materials for a single U.S. data centre from as many as 90 countries, increasing exposure for construction and operations.
Conclusion
Swiss Re’s sigma report frames the hyperscaler-led wave of AI infrastructure investment as a major market development with tangible implications for insurers, capital markets, and the broader economy. Managing the associated accumulation, inflation and supply-chain risks will likely require a mix of risk engineering, alternative risk transfer and financing solutions as industry participants adapt to the changing risk landscape.