Addressing AI risk in underwriting
Tim Hardcastle, chief executive and co-founder of Instanda, has said that reacting to every new artificial intelligence warning is not the same as keeping pace with actual risk. While underwriters understand the technologies, slow updates to policy wordings and rating models often leave insurance products trailing behind the market. Hardcastle noted that insurers should focus on specific, insurable events rather than trying to cover AI as a whole. Examples include models producing incorrect outputs that cause financial loss, systems breaching contracts or regulations, or tools failing required controls. He added that systemic risks, such as an AI system escaping control entirely, are matters for governments and capital markets rather than individual policies.