More than half of UK insurers have now embedded AI into core business functions but a growing gap between ambition and execution is emerging as the industry's most pressing strategic challenge, according to new research.
The findings come from Earnix's Insurance 2026: AI Trends Bulletin, based on responses from 40 UK insurance leaders as part of a wider survey of more than 400 global insurance executives. The report found that while UK insurers have clearly moved beyond the pilot phase, many are struggling to translate isolated AI deployments into enterprise-wide operational decisioning.
According to the research, 55% of UK insurers said AI is already integrated into some business functions. This headline figure marks a significant shift from cautious, proof-of-concept approach that characterised the market as early as 2024. A separate report on Lloyd's market found that around 50% of firms reported limited or no AI implementation in 2025, with adoption accelerating sharply in the 12 months that followed.
That acceleration now has parliamentary weight behind it.
Evidence submitted to the House of Commons Treasury Committee found that more than 75% of UK financial services firms are now using AI, with the largest take-up among insurers and international banks.
Published in January 2026, the Committee's report concluded that the Bank of England, the FCA and the Treasury are exposing the public and the financial system to potentially serious harm through what it described as a wait-and-see approach to AI risk management in financial services.
The Earnix data also showed that UK insurers are concentrating their efforts on operational workflows. Claims processing, policy issuance and the handling of unstructured data are the primary deployment areas, with customer retention and personalisation ranking lower as current priorities. The generative AI figures are particularly striking, with 98% of UK insurers either already using or planning to use generative AI to process unstructured data, a figure significantly ahead of the global average.
Adrian Mincher, head of UK, Ireland and South Africa at Earnix, said the direction of travel is clear but the hard work is only beginning.
"UK insurers have clearly moved beyond experimenting with AI in isolated pilots," said Mincher. "What comes through strongly in this research is that the pressure has shifted to making AI work consistently in the real world, inside pricing, underwriting, claims and customer decision-making, where speed, governance and commercial performance all matter at the same time."
Despite the pace of adoption, Earnix identified a persistent gap between deployment and scale. Thirty percent (30%) of UK insurers admitted they are significantly lagging behind customer expectations on personalisation, and 53% said regulation is moderately slowing AI innovation.
Governance is compounding the problem. Despite strong oversight structures, only 28% of UK insurance leaders strongly agree that their governance cadence is sufficient, highlighting a growing gap between existing frameworks and the pace of AI deployment.
Talent is an equally significant constraint. A separate Accenture survey found that a quarter of insurance executives cited skilled talent shortages as the main factor limiting their ability to extract value from AI.
FDM Group's Mayank Arora noted that many carriers still run on older core systems and have strong actuarial and risk functions but often lack deep AI engineering and data science expertise, with the gap becoming most visible when insurers attempt to move proof-of-concepts into live production systems.
According to the research, 91% of UK insurers plan to increase investment in third-party data, reflecting a recognition that AI performance is only as strong as its inputs. Around 80% of insurers remain concerned about the quality of data feeding their AI models. Meanwhile, a separate industry report found that insurers manage an average of 17 separate data sources feeding premium processes, with data fragmentation identified as the single biggest integration obstacle.
The consensus at BIBA 2026 was that the industry has broadly accepted AI as a commercial necessity but is still working through the legal, regulatory and operational implications, with the firms best positioned being those treating AI adoption as a governance challenge rather than a technology project.
Mincher's conclusion reflects that framing.
"The conversation around AI is becoming far more practical now, less about testing the technology, and more about whether insurers can bridge the execution gap and actually embed AI confidently and at scale into the way decisions get made every day," he said.
