Artificial intelligence has moved from the edges of experimentation to the centre of strategic conversation across the London Market. Virtually every firm is now testing, piloting, or investing in both generative and agentic AI. The pace of change has accelerated, investment is rising, and expectations are growing. However, our recent research into AI adoption across the Market reveals an important truth. Even though enthusiasm for AI is high, true readiness is low. Many firms are moving quickly without the governance, protection, or coordinated strategy that will be necessary for sustainable success.
This report, conducted by Cognizant and sponsored by Microsoft, provides one of the clearest views yet of how London Market firms are approaching AI and where gaps have started to appear. These findings matter because the next two years will determine whether individual firms become leaders in AI adoption or whether they find themselves managing the consequences of early decisions made without enough structure or control.
The purpose of this article is to summarise those findings, provide context for what they mean, and outline the decisions that will separate strategic adopters from reactive ones by 2030.
A Market That Has Fully Embraced AI Experimentation
Almost nine in ten firms in the London Market are either actively investing in AI or running pilot programmes. The Market is therefore not debating whether to adopt AI. It has already crossed that threshold. What varies significantly is the level of maturity with which firms are approaching the challenge.
Investment is set to rise quickly. On average, firms expect to invest around three and a half million pounds in the next twelve months. This rises to eight million pounds in year two and more than fifteen million pounds by year three. A proportion of firms project even higher figures for longer term investment. This trajectory shows how central AI has become to transformation planning.
However, the amount that firms plan to invest is far from consistent. Almost half expect to invest less than one million pounds in the coming year, while only a fifth expect to invest more than five million pounds. This suggests that some firms see AI as a broad strategic shift, while others still view it as a series of contained experiments that may or may not scale.
One of the most important factors behind this divide is the role of the board. Where firms have board approved AI strategies, investment is significantly higher and confidence is almost twice as strong. This demonstrates how leadership alignment influences not just direction but also the organisational belief that the investment will deliver value.
Productivity Is the Primary Driver of AI Adoption
The most consistent driver of AI adoption across the Market is productivity. More than four out of five firms cite productivity gains as a core reason for investing. This reflects the nature of the work the Market undertakes. Underwriters assess lengthy and complex policy wordings. Brokers assemble comprehensive submission packs. Claims teams review extensive evidence and documentation. These activities contain many structured and repeatable tasks that AI can accelerate.
Alongside productivity, firms also reference data quality improvements, cost reduction, and customer experience. These priorities show that firms are taking a practical approach. They see AI as a means to strengthen the work they already do rather than as a mechanism for radical change. However, this practical approach creates a tension. Productivity improvements are likely to prompt cost pressures, especially in firms with private equity owners. Even where AI is framed as a tool for augmentation rather than replacement, productivity gains often lead to expectations of financial savings.
There is also a more fundamental question that firms will need to confront. If AI can significantly accelerate existing processes, should those processes continue in their current form. This is the difference between using AI to make existing work faster and using AI to rethink how the work should be done at all. At present, most firms are still focused on the former.
Three Gaps Are Constraining AI Readiness
Although AI activity is widespread, readiness is limited. Only five percent of firms feel fully prepared to adopt AI at scale. Nearly a quarter believe they are almost ready, while half consider themselves partially ready. More than a fifth acknowledge that they are not ready at all. Three major gaps sit behind these figures.
1. The Governance Gap
Firms understand the importance of governance. More than two thirds believe legal and compliance frameworks are highly relevant for effective AI deployment. Yet only a third have such frameworks in place. Most firms are developing their approach but have not yet reached maturity.
This mismatch between awareness and execution is important. The regulatory landscape is still evolving. Firms must comply with Lloyd’s requirements, FCA expectations, and international regulatory developments. Deploying AI without strong governance creates risk, especially when decision making becomes opaque or when model behaviour is poorly understood. Without clear oversight, even well-intentioned pilots can introduce challenges that become difficult to fix later.
2. The Insurance Gap
One of the most concerning findings is that only thirteen percent of firms have secured liability coverage for their AI deployments. These risks include inaccurate outputs, unintended autonomous behaviour, algorithmic bias, and errors that could cause harm to clients or the firm itself.
It is unusual for a risk management industry to operate with such a high degree of uninsured exposure. The fact that most firms do not have coverage suggests that many have not yet built the governance and documentation that insurers will require. Coverage functions as both a protection mechanism and a readiness indicator. If a firm cannot secure cover, it is likely that its approach is not yet mature enough for large scale deployment.
3. The Coordination Gap
The third gap relates to the pattern of adoption across organisations. A large portion of AI adoption is occurring within individual departments. Underwriting, claims, and IT are experimenting, often independently of one another. Although this promotes innovation, it also creates fragmentation.
When tools and models are developed in isolation, firms accumulate technical debt. Systems become difficult to integrate. Data quality becomes inconsistent across functions. Vendor relationships expand without alignment. All of this increases cost and reduces the likelihood that AI will deliver its full potential. The most valuable benefits of AI rely on connected data and coordinated workflows. Fragmented adoption undermines that ambition.
Two Possible Futures for AI in the London Market
The research indicates that the Market is heading toward a split. Over the next two years, firms will choose one of two paths.
In the first path, organisations build strong governance structures, secure the appropriate insurance coverage, develop a unified strategy, and create the architecture needed for enterprise scale AI. These firms will see genuine productivity improvements, better risk selection, enhanced regulatory credibility, and smoother operating processes. They will be able to rethink how they work, not just accelerate the work they already perform.
In the second path, firms continue to experiment without alignment. Governance remains incomplete. Technical debt grows. Integration becomes more challenging. Regulatory intervention increases. The cost of catching up later rises significantly, and firms risk falling behind competitors that acted more decisively.
The difference between these two futures is not whether firms adopt AI. Nearly all already have. The difference is whether they adopt it strategically.
Four Actions That Will Unlock AI Value
The research points to four actions that should guide firms as they prepare for the next stage of AI adoption.
- Secure board level ownership of the AI strategy.
This creates clear accountability and aligns investment with strategic priorities. - Embed governance from the start.
Governance should be integrated into every pilot and every new initiative. It should never be added later. - Coordinate activity across functions.
Enterprise architecture, data standards, model governance, and vendor management must be aligned. AI cannot succeed at scale if it is built in silos. - Secure insurance coverage.
This provides essential protection and acts as a confirmation that governance and documentation meet the necessary standards.
To explore the full set of findings in greater depth, including detailed data, market analysis, and the complete set of recommendations for London Market leaders, you can download the full research report. The study was conducted by Cognizant and is sponsored by Microsoft. It provides a comprehensive view of the current state of AI adoption and the strategic considerations that will shape success through to 2030. For anyone responsible for defining AI strategy, guiding transformation, or preparing for regulatory change, the full report offers the evidence base needed to make well informed decisions.
