Insights from those on the front line.
How do we ensure the London market is AI-ready? And are we focusing on the right building blocks before we reach for the ‘shiny stuff’?
During a discussion with claims leaders ahead of London Market Claims on October 7th, these questions came as we discussed the practical realities of AI and automation in claims.
One thing was clear: intentionality matters. Whether you're exploring process automation, document ingestion or the potential for large language models to suggest next steps in complex losses, the foundations must be strong. That means structured, standardised data, clear ownership and a shared understanding of what ‘good’ looks like.
Getting over the hype
Despite the buzz around AI, several participants warned against rushing ahead without a purpose. One participant described the challenge of AI experiments springing up across a business – all well-meaning, but disconnected. Without a clear strategy and alignment across functions, organisations risk fragmentation, wasted investment and confused outcomes.
The market is still in early stages. While some firms are using AI to extract key claims data from documents or streamline ingestion, few are yet deploying it for decision-making. Concerns about traceability, hallucination and regulatory scrutiny remain high. But most agree that progress is inevitable and that hesitation brings its own risks.
So how do we get there?
Many pointed to data quality and consistency as the critical first step. If your claims data today is 80% incomplete or inconsistent, AI won’t solve that – it will amplify it. Several firms are investing in automation tools specifically to improve foundational data, often starting with high-impact points such as cause of loss or policy identifiers. Others are testing AI to support claims handlers by surfacing relevant information or guiding next steps, but always with human oversight.
There was also discussion about skills and the changing role of claims professionals. One participant described the shift as “liberation from admin”, freeing up time to focus on insights, client relationships and complex advocacy. AI should support, not replace. But success depends on trust, both in the tools and among the teams using them.
Importantly, it was emphasised that claims data is not just for claims. Much of the value sits with internal stakeholders in pricing, underwriting and risk. To secure investment, claims leaders must demonstrate not just operational gains, but wider business value.
What will be the market’s tipping point for AI? What will finally tip us from experimentation to real transformation? And how do we ensure we don’t miss that inflection point?
To explore these questions with your peers, access new thinking and shape the market’s approach to AI, join us at London Market Claims on October 7th.
For more information about the events & membership, contact Tom@TIN.events.
Sign Up to TINsights
Where we share our latest blogs, industry reports and insights
