Allianz warns AI losses are being buried in existing insurance claims
The insurer is tracking multimillion-euro claims, while unclear policy language leaves businesses uncertain about which coverage would respond.
Loading page…
The insurer is tracking multimillion-euro claims, while unclear policy language leaves businesses uncertain about which coverage would respond.
Listen to this story
Allianz Commercial says AI-related losses are emerging inside ordinary cyber, directors’ and officers’, product-liability and crime claims, making current AI claim totals an incomplete measure of exposure. The insurer is tracking claims involving millions of euros, while estimating that some observed cases could each produce about $100 million in losses across the insurance industry—a potential-loss estimate, not a confirmed payout. For businesses, the risk is both whether policies cover AI-related harm and whether operations can withstand outages or failures in shared AI systems.
Allianz groups AI losses into direct harm, AI-related governance or reputational claims, AI-accelerated malicious activity, and losses worsened by business dependence on AI.
The D&O Diary counted 22 AI-related shareholder filings through August 2026, compared with 16 for all of 2025 and seven in 2023, according to cited figures.
A 2026 RAND report found few insurers explicitly cover AI losses; exclusions are growing, while most policies remain silent and leave coverage dependent on how a claim is characterized.
Businesses may already be suffering AI-related losses without those losses appearing in AI claims totals. Allianz Commercial warns that claims recorded under cyber, executive liability, product liability and crime insurance can obscure the technology’s role. Its report, covered by Insurance Business, argues that modest confirmed losses give an incomplete picture of business exposure.
The insurer says it is tracking AI-related claims involving millions of euros. Individual cases it has observed could each generate total insurance-industry losses of around $100 million. That larger figure is an estimate of potential losses across insurers, not a confirmed payout. Allianz’s warning concerns both losses already emerging and exposure that current claims classifications may fail to capture.
Allianz separates losses into four categories according to how AI contributes. The distinction spans a system making a damaging decision, people misrepresenting its capabilities, criminals using it and businesses becoming dependent on it.
The dependency category does not require an AI malfunction. Allianz describes a large retailer whose AI-dependent supply chain fell back to paper-based processes after a third-party technology outage. Inventory imbalances, delays and significant financial losses followed. Michael Daum, who heads Allianz Commercial’s AI Risk and Claims Center of Excellence, identified the reliance on AI-supported processes and limited ability to operate without them as the primary loss driver.
One source of executive-liability exposure is AI washing: overstating or misrepresenting AI capabilities in investor disclosures. Allianz reports confirmed notifications under directors’ and officers’ liability insurance, commonly called D&O, involving this issue. The claims concern what companies and their leaders told investors, rather than necessarily arising from a defective AI output. Allegations about corporate conduct can therefore bring AI exposure into an existing policy.
Coverage is not uniform. A 2026 RAND Corporation report cited by Insurance Business found that a minority of insurers explicitly cover AI losses, a growing number are adding broad exclusions, and most remain silent. Where a policy says nothing about AI, its response depends on how a loss is characterized, with differences across insurers and insurance lines. Silence does not give a business a clear answer about whether a particular AI-related claim will be covered.
Allianz also identifies a risk that extends beyond any one policyholder. Businesses use common large language models, training data and cloud infrastructure. A flaw or outage affecting those shared systems could produce losses at many companies at once. For insurers, that creates the possibility of multiple claims arriving from the same underlying failure, rather than isolated incidents at unrelated businesses.
AI can also reduce losses. Allianz’s claims analysis finds materially lower breach costs among organizations using AI-driven detection tools. As Beinsure describes the report, CEO Thomas Lillelund urges businesses to establish clear accountability, controls, human oversight and fallback arrangements. The practical task is not simply to identify where AI is used, but to preserve the ability to respond when those processes fail.
Loading discussion...
Join the conversation
Explain whether explicit wording would offer meaningful protection or just more exclusions.
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.