Artificial intelligence (AI) is creating new forms of liability for organisations and challenging traditional insurance frameworks. Where an AI system generates inaccurate advice, produces misleading outputs or makes defective automated decisions, the resulting loss may contain elements of cyber risk, professional negligence and product failure.
Blurring of traditional insurance boundaries
As organisations increasingly integrate AI into customer-facing products and professional services, insurers and businesses are grappling with a fundamental question: which insurance policy should come into play? Historically cyber insurance, professional indemnity insurance and product liability insurance have responded to distinct categories of risk. Cyber insurance addresses network security failures, data breaches and cybercrime, typically involving external threat actors. Professional indemnity insurance responds to negligent advice or professional services, while product liability insurance addresses harm caused by defective products.
AI does not fit neatly within these established categories. An AI system may generate inaccurate advice, produce defective outputs or make automated decisions that cause financial loss without any involvement of an external threat actor, human error or a conventional product defect. In many cases, the AI may be operating exactly as designed.
As a result, AI-related losses often exhibit characteristics of multiple risk categories, creating uncertainty as to whether they constitute a cyber event, a professional services error or a product failure. This uncertainty can have significant coverage implications where multiple policies may potentially respond, or where insurers contend that a claim falls more appropriately within the scope of another insurance product.
Evolving and new product development
Insurers are responding to these challenges in two ways. First, many are refining policy wording to clarify the allocation of AI-related risks among cyber, professional indemnity, technology errors and omissions (Tech E&O) and product liability policies. Definitions of ‘professional services’, ‘technology products’ and ‘cyber events’ are increasingly being drafted with AI use cases in mind, while endorsements are being introduced to address liabilities arising from AI-generated outputs and automated decision-making.
Second, the market is beginning to develop hybrid products that combine elements of cyber, Tech E&O and professional indemnity cover. These policies are designed to respond to losses that fall between traditional insurance categories, such as claims arising from AI-generated advice, defective algorithmic outputs or failures in AI-enabled services. Some insurers are also exploring standalone AI liability products aimed at organisations that develop, deploy or commercialise AI systems. While these products remain in their infancy, they reflect a broader recognition that traditional insurance silos may be ill-suited to the complex liabilities created by AI.
Implications for organisations
For organisations deploying AI, the key challenge is no longer simply obtaining insurance coverage. It is understanding how different policies interact and whether AI-related liabilities are adequately addressed across the organisation’s risk transfer programme
Risk and legal teams should review policy wordings carefully, identify potential overlaps and gaps between cyber, professional indemnity and product liability cover, and engage with their brokers and risk advisors regarding AI-specific exposures. As AI adoption accelerates, organisations may find that the critical question is which policy responds when AI related losses arise.
