AI is already being used across the London Market, but the more difficult question now is how far firms have moved it beyond experimentation and into day-to-day activity.
A 2025 Lloyd’s Market Association survey of 81 London Market firms found that 40% were actively using AI in some areas or had integrated it more widely into workflows, while another 47% were experimenting with AI tools. Furthermore, 65% had yet to deploy or experiment with agentic or generative AI within underwriting or claims.
Clearly, the Market has underwriters testing AI-assisted workflows, firms using technology to extract submission data and growing interest in algorithmic underwriting. Lloyd’s also continues to bring technology businesses directly into the Market through the Lloyd’s Lab, with its September 2026 cohort placing operational efficiency among its areas of focus. Yet much of the activity is concentrated in particular teams, processes and use cases. As a result, progress is happening at varying speeds across different firms.
So is the London Market ahead or behind on AI? It depends on whether we are measuring interest and experimentation, or the ability to use AI consistently across the business.
The wider UK insurance sector compares strongly with other areas of financial services.
The Bank of England and FCA’s most recent published AI survey found that 95% of insurance respondents were already using AI, the highest proportion of any sector surveyed. General insurance also accounted for around 10% of the AI use cases reported across financial services, with high-materiality use cases particularly prevalent in general insurance, risk and compliance.
As for the London Market, submission intake and data extraction are obvious areas for AI because insurers continue to receive large volumes of unstructured information. In Guidewire’s 2026 survey of more than 250 brokers dealing primarily with London Market insurers, 42% identified automating submission intake and data extraction as a leading AI use case, while 38% highlighted exposure management.
In the same research, 51% of brokers said the shift towards algorithmic or fully digital underwriting was already happening. While these changes are indeed meaningful to the way risk is processed and assessed, they also show why headline adoption numbers only tell part of the story.
Research from Cognizant, presented by London Market Forums in December 2025 and drawing on insights from more than 100 senior leaders across 72 organisations, found strong interest and experimentation alongside a lack of clear enterprise-wide strategy in many organisations. Governance, data quality and accountability were among the issues limiting wider adoption.
Similar conversations have continued through 2026; London Market underwriting leaders have described a move away from early experimentation towards practical deployment, although there remains a gap between having AI in selected workflows and demonstrating measurable value at scale.
The challenge now for leadership teams is deciding which tools should become part of the operating model, integrating them with existing systems and demonstrating that they improve underwriting, claims or operational performance requires much more work.
AI makes existing data problems harder to ignore, particularly for London Market firms, where teams work with complex information moving between brokers, underwriters, claims teams and third parties, often across systems that were built at different times and for different purposes.
The LMA’s 2025 AI research identified data quality and availability among the primary barriers to adoption. More recent broker research found that 37% of brokers saw fragmented standards as a leading cause of poor data quality.Among respondents moving ahead with their own technology strategies, 31% also expressed concern about integration because of legacy core constraints.
Following the move away from Blueprint Two in March 2026, Lloyd’s has placed renewed emphasis on process simplification, common data standards and incremental technology modernisation - foundations which directly affect what is possible.
In practice, an underwriting tool cannot reliably analyse information that arrives inconsistently. Similarly, automation cannot remove much friction if people still have to reconcile information between systems. AI-generated analysis becomes harder to trust if nobody is clear on where the underlying data came from or who owns it.
Investment in AI therefore needs to sit alongside the less exciting work of improving data, processes and integration.
Some parts of the Market are moving faster than others.
For example, underwriting has attracted considerable attention, particularly around submission handling, exposure analysis and algorithmic approaches. Meanwhile, the attention on claims is more mixed.
Guidewire’s broker research found that senior respondents were more likely to prioritise AI and automation for submission intake and market search than for claims management. Only 19% identified claims management as a priority area for AI investment. The same research pointed to continued frustration with repetitive processes and settlement times.
There is activity within claims, and London Market claims leaders have been exploring AI, automation and digitalisation in greater depth. Deloitte’s research with 22 London Market Chief Claims Officers described transformation activity as accelerating, driven in part by AI opportunities and wider market modernisation.
As such, firms will struggle to realise the full value of AI if improvements remain concentrated around the front end of the insurance lifecycle. Information produced during claims should feed back into underwriting and portfolio decisions, while operational improvements after placement directly affect broker and client experience.
As firms move AI closer to underwriting and other material decisions, governance becomes more important.
The regulatory direction in the UK currently gives firms room to innovate; the FCA has said it does not plan to introduce a separate set of AI-specific rules, instead relying on existing frameworks and expectations around governance and controls.
The PRA is also monitoring increasing AI use within regulated firms and has committed to continuing its work with industry on responsible adoption during 2026/27.
The practical questions for London Market insurers are already here; firms need to know where AI is being used, who owns each use case, what data it relies on and how output is reviewed. The level of control should reflect the significance of the activity, particularly as AI moves from administrative work into areas that influence risk selection, pricing, reserving or claims decisions.
Third-party reliance also needs attention, as a significant proportion of AI capability will come through external technology providers, creating dependencies that need to sit within existing operational resilience and outsourcing frameworks.
On experimentation and adoption, the Market has moved a long way. AI is already being used, algorithmic underwriting is growing and conversations have shifted towards practical use cases and measurable value. However, consistent implementation at scale remains less developed across much of the Market.
Data remains fragmented in parts of the Market, and legacy systems still restrict integration. Firms are at very different stages of adoption and some areas of the insurance lifecycle have attracted considerably more investment than others.
The next phase will therefore be less about proving that AI has a place in insurance. Firms now need to decide where it improves performance enough to justify investment, and build the data, controls and operating model around those use cases.
For some London Market businesses, that work is already well underway, but others still need to understand what AI is currently being used for inside their organisation before deciding what comes next.
Brighter Consultancy works with London Market firms across finance, risk, compliance and transformation, helping organisations understand how new technology affects their operating model, governance and wider change priorities.
If you are reviewing how AI should be used across your organisation, speak to our team to discuss where to start.