Lockton Re and Armilla warn AI risks sit outside traditional insurance frameworks

Lockton Re, working with Lockton International and Armilla AI, has published a report arguing that artificial intelligence is reshaping insured risk in ways that existing commercial insurance structures were not designed to address.

The report, titled “Ready or Not: The Impact of Artificial Intelligence on Insurance Risks”, examines how the probabilistic nature of AI systems challenges traditional coverage triggers across commercial lines.

Oliver Brew, co-author and head of cyber centre of excellence at Lockton Re, said AI introduces a fundamentally different risk profile to conventional software.

“The benefits of AI are clear. It’s also important to recognise that AI is fundamentally different to traditional software,” Brew said.

“The probabilistic nature of AI and its ability to synthesize massive amounts of data mean it is prone to errors, which will have huge implications wherever implemented,” he added.

“There are no sectors of the economy that are insulated from the potential impact of AI,” Brew said.

“As an industry we need to prepare for how these rapidly evolving risks are underwritten across commercial insurance and what emerging claims patterns will look like,” he said.

“The current lexicon and frameworks for insurance products and risk categories were not designed with these systems in mind and are increasingly misaligned with how AI-related losses occur,” Brew added.

The report maps AI related exposures across multiple commercial classes and highlights areas where coverage may be silent, fragmented or misaligned with how AI failures present in practice.

It also examines systemic risk arising from shared AI infrastructure, common foundation models and correlated behaviour across portfolios.

Baiju Devani, co-author, chief technology officer and co-founder at Armilla AI, emphasised that AI introduces novel perils that underwriting tools are not yet equipped to price.

“The underwriting of AI risk needs to consider the novel perils created,” Devani said.

“Even if we recognise the peril, there is a growing gap between what insurers intend to cover and what they actually cover,” he added.

“The underwriting toolkit needs to develop at pace to model and price AI risks,” Devani continued.

“Additionally, the challenge for the insurance industry is not whether AI will create systemic risk events, but when, and if underwriting practices can keep pace,” he added.

The report includes a claim scenario involving an AI powered customer service chatbot that generates incorrect warranty commitments without any system breach or malicious activity.

The example illustrates how AI driven losses can fall outside traditional cyber and liability triggers while still creating material financial and regulatory exposure.

Karthik Ramakrishnan, CEO and co-founder of Armilla AI, said ambiguity in policy wording risks undermining the value of insurance cover.

“The report highlights how ambiguity in insurance contracts reduces the perceived value of coverage as AI systems move into core business decision-making,” Ramakrishnan said.

“The silence in policy language creates uncertainty when claims arise,” he added.

“Both insurers and policyholders benefit from greater clarity in how AI-related risks are addressed to support responsible adoption of the technology,” Ramakrishnan said.

The report argues that AI should increasingly be considered as its own category of risk classification within insurance frameworks.

Read the report here.