Three themes rewriting the rules of insurance
AI infrastructure investment, capital market fragmentation and supply shocks are reshaping insurance discussions as brokers help clients protect higher-value assets from evolving, interconnected risks. But the most tangible pressure point remains inflation, with higher-for-longer energy prices once again forcing economists to rethink key macroeconomic assumptions.
Conflict demands macro forecast revisions
Locally, businesses and consumers saw a sharp upward move in prices confirmed in the May and June 2026 CPI numbers published by Statistics South Africa at 4.0% and 4.4% respectively. And globally, rising energy prices and supply chain constraints have forced a revision of inflation and economic growth numbers. At the media launch of World Insurance SIGMA 2/2026, global reinsurer Swiss Re pencilled in a 0.3 percentage point reduction in world GDP growth to 2.5% alongside an almost one percentage point rise in global CPI inflation.
“For insurance markets, this definitely will mean higher claims inflation,” said Jérôme Haegeli, Group Chief Economist and Head of the Swiss Re Institute. The chief economist said that artificial intelligence (AI) was here to stay, and that global risks were evolving in response to the race for digital and natural resources. He noted that the insurance and reinsurance industries would be called on to protect businesses from more frequent supply shocks and tail risks. Although the reinsurer did not expect runaway inflation, it warned that price shocks often took six months to a year to transmit to the property and casualty insurance market.
“We expect supply disruptions to have an impact on claims inflation; and property prices and construction costs are likely to see a more sizeable impact than on headline inflation,” Haegeli said. World Insurance SIGMA 2/2026 points to a fragmented primary insurance market in which non-life premium growth softens or bottoms out and life premiums remain resilient. The outlook is better for emerging markets ex-China; but the US and other advanced economies could see real premium growth fall close to zero, a two-decade low.
Profit metrics remain robust
The good news for insurance brokers and their business and household clients is that primary insurers and reinsurers remain profitable. SIGMA 2/2026 suggests the return on equity across primary non-life insurers will slow from 14% to 11%, still well above the cost of capital. More broadly, approaches to non-life underwriting will be influenced by three themes including supply shocks; fragmentation in capital markets; and AI-driven infrastructure, each with its own set of opportunities and challenges.
The reinsurer warned that the contributing factors to supply chain shocks are complex and plentiful. At the geopolitical level, businesses have to deal with the consequences of conflict, energy prices, industrial policy and nearshoring. Insurers and reinsurers also face mounting cyber risks arising from the development and adoption of AI and technology. Finally, the traditional exposures to natural catastrophes and pandemics remain in play. Aside from inflation and higher interest rates, this theme elevates risk across the insurance landscape.
Are contract wordings context-fit?
Haegeli said that the insurance industry would have to revisit contract wordings and clauses in the context of accumulation risk and extreme uncertainty. Accumulation risk refers to the concentration of insured exposures that could give rise to multiple claims from a single event or common underlying cause. As an example, consider a manufacturer that deploys the same AI system across several plants. A flaw in that system could trigger losses under multiple policy sections, across multiple sites at once. And this from a risk source that hardly featured in underwriting conversations five years ago.
World Insurance SIGMA 2/2026 contends that fragmentation in capital markets is a greater risk to the global economy and insurance markets than fragmentation in goods trade. “The ability to move capital around by financial accounts and financial flows has become much more volatile, much stickier,” Haegeli said. Concerns centre around national economic security policies making capital less mobile, thereby reducing the diversification benefit of spreading risk across regions and lines of business, and pushing up the cost of capacity just as the industry needs capital to absorb larger shocks.
AI-related infrastructure boom
The third theme, being the challenges and opportunities arising from AI-driven infrastructure investment, was framed in the ‘shock and awe’ capex numbers being bandied about by the so-called hyperscalers. To illustrate the sheer scale of investment, a report by Goldman Sachs holds that cumulative AI-related capex could top USD7.6 trillion between 2026 and 2031.
Haegeli said these forecasts were good for economic growth, productivity growth and, of course, for insurance premium growth. However, he warned of accumulation risk in the context of the geographic footprint of US-based data centres and natural catastrophe exposures. “Fact is, many of the data centres in the US are concentrated in high-risk locations; the Swiss Re CAT team estimates that almost half of US data centres are likely to sit in significant to very high tornado zones, meaning that underwriting of these data centres is also at risk,” he said. The assets on cover at a large data centre can easily reach USD20 billion.
The evolution of risk management
The stage was set for Ivan Gonzalez, CEO at Swiss Re Corporate Solutions, to comment on practical responses to broader themes. Gonzalez commented on the evolution of risk management at Fortune 500 firms, from “one or maybe two people sitting somewhere at a relatively lower level within the organisation” to larger teams that considered “the overall resilience of the enterprise”.
In the old days, you had someone who telephoned a broker to buy insurance; nowadays, the risk management team, working with an insurance broker, must achieve “a holistic view across loss prevention, insurance and risk financing”. “Building resilience requires a strong collaboration across all stakeholders,” Gonzalez said. There were some fascinating estimates around how the AI data centre buildout might transmit to the insurance and reinsurance sectors, with Swiss Re estimating up to USD90 billion in non-life insurance premium, on an aggregate basis, through to 2030.
To support this level of protection, the industry would have to ‘find’ in the vicinity of USD400 billion in capital. “If you look at the risk around these data centres [you can conclude that] traditional underwriting is no longer sufficient,” Gonzalez said. To properly insure this type of infrastructure, risk managers will have to consider natural catastrophe exposure alongside myriad interconnected risks such as electricity networks, sophisticated cooling systems and cyber risks introduced by connected operational technologies, to name a few.
In the unthinkable scenario of a total loss, the rebuild of such a facility would take years and require coordination between suppliers based in over 90 countries.
Coordinated risk transfer responses
Gonzalez closed by arguing that data-centre risk calls for more than traditional insurance capacity. Risk engineering has to sit alongside alternative risk transfer and the credit and surety support required through construction, financing and operation. “All the resilience that is required comes predominantly through the ability of all stakeholders to partner, to look at this [risk] holistically,” he concluded. It was a fitting close to a presentation that framed insurance and reinsurance as the shock absorbers that allow economic development to continue.
Writer’s thoughts:
Global non-life insurers are being challenged to protect assets whose value can exceed SA’s entire annual non-life insurance premium pool. Is this something that keeps you awake at night, or do your insurance worries sit closer to home? Please comment below, interact with us on X at @fanews_online or email us your thoughts [email protected].
