UK consumers want human checks in insurance AI use

Guidewire has released its latest European consumer survey on attitudes to artificial intelligence in insurance. The UK findings show support for AI use when human checks remain in place.

The study found that 30% of UK customers would be happy for their insurer to use AI tools to make decisions about policy pricing. Acceptance was higher in tasks where respondents could see a direct benefit: 38% were comfortable with AI helping them complete insurance documents and policy applications, and 39% were comfortable with AI supporting human call handlers in answering questions.

The results were broadly steady from the previous year, suggesting familiarity alone has not yet produced a wider shift in public confidence. Instead, respondents set out clear conditions under which they would be more willing to accept AI in insurance processes.

Human intervention was the most frequently cited requirement, named by 33% of UK consumers. Transparency followed at 26%, while 23% said third-party regulation was needed to keep the technology in check.

The survey also showed a sharp divide between the wider public and people who already use AI tools regularly in their own lives. Among daily AI users, 63% said they were comfortable with "human-free" policy decisions, compared with 30% of the broader UK population.

That pattern extended across other parts of the insurance process. Four in five daily AI users said they would be happy for the technology to help fill out insurance documents and policy applications, compared with 38% of UK consumers overall. Nearly 60% said they were comfortable with AI deciding and processing claims or determining claim value, against 27% of the wider public.

The findings suggest insurers face a mixed public mood as they expand AI use in pricing, service and claims. Customers appear more willing to accept automation when it reduces administrative effort or improves speed, but remain more cautious when it affects decisions with direct financial consequences.

Trust question

The research covered 4,004 consumers across the UK, France, Germany and Spain who had bought or renewed a general insurance product, or made a claim, in the previous 12 months. Of those, 1,000 respondents were from the UK.

The survey has tracked customer attitudes since 2020. The latest UK results suggest insurers are dealing with a trust issue shaped not just by the technology itself, but by how visible and accountable its use is to the customer.

Even among frequent AI users, demand for oversight remains significant. The survey found that 39% of daily users still wanted a human kept in the loop, while 30.6% pointed to transparency as a condition for trust.

That may matter for insurers looking to move AI beyond administrative support into underwriting, pricing and claims decisions. Public willingness appears strongest when a person remains involved and when customers understand what the system is doing.

Charles Clarke, Group Vice President at Guidewire, said: "AI is playing an increasingly important role in the insurance industry, and customers are becoming more comfortable with its use. Our report shows that when customers clearly see its value, they are significantly more likely to accept AI within the insurance process.

"To further enhance acceptance, customers are calling for greater transparency, regulation, and human oversight. Insurers should work collaboratively with one another, with technology providers, and with regulators to meet these expectations and build lasting trust in how AI is used whether those capabilities come from their core platforms, embedded assistants, or broader AI partners."

The figures come as insurers across Europe examine how far they can automate parts of customer service and decision-making without provoking regulatory or consumer backlash. In the UK market, where price sensitivity and service standards remain under close scrutiny, the balance between efficiency and accountability is likely to shape how quickly AI is adopted in customer-facing roles.

Michael Cook, Partner at PwC, said: "The findings underscore a clear desire among UK customers for a balanced approach to AI in insurance, embracing the efficiencies and convenience it offers while ensuring that human judgement remains integral to decision-making, especially as AI moves beyond efficiency gains into more value-add work and, ultimately, different ways of operating.

"As AI becomes more embedded in daily life, insurers must prioritise transparency and robust regulation to maintain consumer trust and confidence, including adopting appropriate governance and frameworks and exploring the role of AI to 'manage AI'. Striking this balance will be key to unlocking AI's full potential in delivering fairer, more personalised insurance services and moving to a very different way of operating with a combined people and agent workforce."