Munich. According to Thomas Lillelund, CEO of Allianz Commercial, artificial intelligence (AI) could create risks that “cannot be fully insured”. Public-private solutions or partnerships are necessary for extensive major losses that affect many customers at once, says Lillelund, who has been head of Allianz's industrial insurance business for a good year and a half, in an interview with Handelsblatt.
Lillelund is particularly concerned about companies' growing dependence on a few providers of large AI models. “This dependency could increase,” warns Lillelund. If one of these providers fails or limits its services, numerous companies could be affected - and, for example, suddenly no longer be able to implement important business processes. Risks that affect many companies at the same time are called cumulative risks.
The topic also concerns representatives of other insurers. Clemens Jungsthöfel, CEO of the world's third largest reinsurer Hannover Re, warned in an interview with the Handelsblatt about the economic dependence on digital infrastructures due to the increasing use of AI. Premiums do not reflect these risks in reinsurance, he said.
“As insurers, we have to think about how we can protect against such accumulation risks,” says Lillelund. At the same time, society must ask itself whether it even wants such a concentration. Allianz Commercial has founded its own center to investigate AI damage and identify possible accumulation risks.
Under the Allianz Commercial brand, Allianz combines the business of the subsidiary Allianz Global Corporate & Specialty SE (AGCS) for large corporate customers and the business of Allianz's operating units, which insure medium-sized companies in various countries. In 2025, Allianz Commercial generated global gross premium income of around 17.3 billion euros.
From Lillelund’s perspective, the risks posed by AI and the potential frequency and severity of AI-related harm are increasing. “With AI we see many parallels to developments in the cyber sector, especially when it comes to so-called silent damage,” explains Lillelund.
The insurance industry would have to pay for damages for which the contracts were not originally intended. Lillelund continued: "AI risks now permeate many other products and must be underwritten more transparently. I think it is possible that AI risks will become a separate insurance line in the future."
Specifically, AI changes the risk profile of companies in three ways, explains Lillelund. First, it could cause immediate damage – for example, if an AI-supported system causes physical damage or direct injury. According to Lillelund, Allianz Commercial is still seeing few claims reports in this area. However, they are likely to increase in the future.
Secondly, AI could indirectly cause damage. Lillelund cites legal disputes as an example, for example when a company presents its own AI capabilities as better than they actually are - for example as part of an IPO.
AI is just one of several significant risks for Allianz Commercial’s customers. Climate change, cyber attacks, supply chain disruptions and business interruptions also played an important role. “Geopolitical tensions are the new normal, but at the same time the complexity of the risks is increasing,” says Lillelund.
A current analysis by the reinsurer Swiss Re and the London School of Economics also comes to the conclusion that risks are more closely linked today than they were a few years ago. This would create new avenues through which initially limited disruptions could spread faster and further throughout the economy. AI and supply chains developed into important hubs.
Ivan Gonzalez, head of Swiss Re Corporate Solutions, says: “A company can appear well diversified until it becomes apparent that suppliers, technology and customers access the same infrastructure.” A single disturbance could then affect several areas at the same time. “If you know these dependencies, you can reduce risk concentrations and make better decisions about which risks should be borne yourself and which should be transferred,” Gonzales continued.
Meanwhile, Allianz Commercial itself is also using artificial intelligence to accelerate its own processes. Using a tool to review environmental, social, governance and reputational risks, the company has reduced insurance policy processing time from 20 minutes to two minutes, Lillelund says. The response time for an initial assessment of complex cases has also fallen from five days to one day.
For Lillelund, increased use of AI within the company is also important so that employees are given more freedom to implement Allianz Commercial's growth strategy. Above all, Lillelund wants to promote international expansion.
“We have a significant market position in Germany and do a little more than a fifth of our business there,” he explains. In the USA, for example, the largest insurance market in the world, the market share is still relatively small. “We want to make further progress, especially in business with particularly complex special risks.”
