Newsroom

The New Insurance Customer Is an AI Agent. It Wants Proof, Not Just Price.

Accenture's latest Consumer Pulse Research, drawn from more than 25,000 consumers across 16 countries, makes a claim that should reshape how insurers think about distribution: more than 70% of people are expected to be influenced by AI in their home and motor insurance decisions within the next 12 months.

Newsroom 18 June 2026

Accenture's latest Consumer Pulse Research, drawn from more than 25,000 consumers across 16 countries, makes a claim that should reshape how insurers think about distribution: more than 70% of people are expected to be influenced by AI in their home and motor insurance decisions within the next 12 months. Habitual renewal, long the industry's quiet safety net, is weakening fast. Among Gen Z and Millennials, 86% already expect AI to shape at least half of their insurance shopping.

The detail that matters most is not the adoption curve.

It is what these AI agents are actually optimising for. Accenture's data shows that consumers are not simply asking AI to find the cheapest policy. Around 46% say they trust AI recommendations specifically because they expect those tools to surface product quality and claims performance alongside price. The agent acting on a customer's behalf wants to know whether a claim will actually pay out as promised, not just whether the premium is competitive.

That requirement runs straight into a structural problem. Most insurance products and claims histories are not built to be read by machines. They exist as dense policy wording, scattered documentation and claims data that is rarely structured in a way that allows confident, verifiable comparison. An AI agent trying to assess claims reliability across providers is, in most cases, working with very little to go on.

Accenture's own industry lead made the implication explicit, stating plainly that the insurers who win in an AI-mediated market will be those whose propositions are clear, verifiable and genuinely competitive, which means making products machine-readable and claims verifiable, and ensuring visibility wherever agents compare and decide.

There is a second finding worth sitting with.

Only 8% of consumers are comfortable with full end-to-end automation of their insurance decisions without human oversight, while 42% want to review options themselves before anything is decided. Even as AI takes on more of the discovery and comparison work, people still want to see the reasoning, not just receive a recommendation.

Put those two findings together and the requirement becomes clear. Insurers need infrastructure that can make claims history and product detail verifiable to an AI agent, while still being transparent and explainable to the human who ultimately wants to see why. That is not a marketing or product description challenge. It is a data architecture challenge, and it is the one we are built to solve.