AI as a challenge for insurability

AI is now simply part of everyday life: as an assistant in office processes, as an optimization machine in logistics and production, or as a building block in software products. However, the more AI influences core processes and the more autonomously it prepares or makes decisions, the more often a promise of productivity turns into a risk area that can trigger classic damage scenarios. "Property damage, business interruption, liability claims, cyber incidents, or fraud are just a few examples of how broad the risk area of AI can be," says Alexander Skorna, Managing Director of Funk Consulting. "The risk depends less on the buzzword AI and more on where it is used, how high the degree of automation is, and which control mechanisms, such as internal approvals, verifications, or plausibility checks, are actually effective." 

External studies confirm this development. In the Allianz Risk Barometer 2026, AI is listed as a global business risk for the first time and ranks second. The World Economic Forum's Global Risks Report 2026 also shows that risks related to information integrity and technology are becoming increasingly important. Misinformation and disinformation are among the most serious risks.

"Many AI risks do not materialize as new types of damage, but as familiar events with unusual origins," says Nadine Bewarder, Head of Funk Market Management. When AI is used in production or operational processes, misinterpretation of sensor data or framework conditions can lead to control errors. An example from industrial property insurance illustrates the emergence of new chains of causality: If an AI system controls valves for flammable liquids and incorrectly assesses a situation as non-critical, this can lead to leakage, ignition, and subsequently fire or explosion damage. For the insurance contract, the damage initially looks like classic fire damage—only the causality is more complex.

The mechanism is similar in logistics and transport. AI-optimized route planning and tight time windows can increase operational pressure, making operating errors during loading and unloading more likely. In temperature-controlled supply chains, errors in AI-supported monitoring can lead to spoilage of goods. "In addition, there are new interface risks – for example, from autonomous loading and unloading systems or system failures that cause downtime and delivery delays," says Alexander Skorna. At the same time, the risk of manipulation is increasing, for example through fake freight carriers or AI-supported social engineering attacks.

Studies and practical experience point in the same direction: AI will become more deeply integrated into the core processes of the value chain in the coming years. This will increase the relevance of resilience, liability issues, and clean insurability. The solution is rarely "more insurance at any price." Instead, companies should create transparency about where AI is actually used in their own business, what scenarios may arise from this, and what coverage and contract details these scenarios entail.

We would be happy to engage in a technical risk dialogue with you to analyse and evaluate your individual value chain in a structured manner. We accompany you from the classification of your AI applications and their degree of automation, through governance and controls, to coverage interfaces and optimization options in your existing insurance portfolio. Together, we will find a suitable solution—please feel free to contact us.