2026-07-20 · Source: Insurance News Australia
Summary in 3 Points • AI agent economy projected to handle trillions by 2030, posing unpriced risks for insurers • Insurers' exposure largely silent within existing cyber, professional, and general liability policies • Report urges industry collaboration on AI risk management infrastructure --- A new report from the **Artificial Intelligence Underwriting Company** details the risk posed by the rapid development of AI technologies, particularly agent models capable of autonomous actions. These models, projected to handle trillions of dollars in transactions by 2030, present largely unpriced and invisible risks within existing insurance policies, including cyber, professional, and general liability. The report, supported by research from **Anthropic**, **OpenAI**, **QBE**, and **Aon**, warns of potential correlated losses due to the dominance of a few AI models, such as **OpenAI's ChatGPT**, **Anthropic's Claude**, and **Google's Gemini**. It suggests that insurers risk repeating past mistakes seen in cyber insurance unless they proactively manage these emerging exposures. For the **London Insurance Market**, the emergence of the AI agent economy directly impacts cyber and professional liability lines, where many **Lloyd's syndicates** have significant exposure. The call for industry collaboration to build shared infrastructure for AI risk management is particularly relevant, as London is a global hub for specialty insurance and reinsurance. The potential for correlated losses due to the concentration of AI models calls for advanced catastrophe modelling and risk selection processes, areas where the London Market traditionally excels. **Underwriters** and **brokers** should prioritise understanding the effects of AI agent models on existing policies, particularly regarding silent cyber exposures. Developing expertise in AI-related risks and collaborating on industry-wide standards and incident data collection will be crucial. **Risk managers** must assess their organisations' reliance on AI technologies and ensure adequate coverage for both first-party operational losses and third-party legal liabilities, adapting their strategies to address the unique challenges posed by the AI agent economy.
For the London Insurance Market, the emergence of the AI agent economy directly impacts cyber and professional liability lines, where many Lloyd's syndicates have significant exposure. The call for industry collaboration to build shared infrastructure for AI risk management is particularly relevant, as London is a global hub for specialty insurance and reinsurance. The potential for correlated losses due to the concentration of AI models calls for advanced catastrophe modelling and risk selection processes, areas where the London Market traditionally excels.