But KPMG warns that ethical and governance issues are core implementation challenges. The report highlights “ethical challenges” as the single largest obstacle to AI adoption, and urges insurers to build model inventories and explainability frameworks so AI deployment preserves customer trust. As Jacques Cornic, KPMG’s EMA lead for insurance, notes in the report, “To get the most from AI, it’s important to identify how your AI models could create risks for policyholders or employees. This calls for an inventory of models that use AI, and a clear description of how you use data, to gain trust in the output.”