Claims transformation is about more than introducing new technology. For the London market, the opportunity lies in combining AI, better data and specialist expertise to strengthen decision-making, improve client service and bring claims insight closer to underwriting.
Discussions at The Insurance Network’s London Market Claims conference highlighted a practical challenge: how can firms move beyond experimentation and deliver lasting improvements?
The pre-event survey revealed both confidence and a readiness gap. While 60% of respondents were confident about scaling AI beyond pilots, only 19% were very confident. Meanwhile, 54% identified operating model redesign as a high priority. Although 63% felt culturally prepared for change, just 11% said it was embedded.
These findings point to a market that recognises the opportunity but still has work to do to turn ambition into everyday practice.
Start with the operating model
Automating an inefficient process can make it faster without making it better. Before selecting technology, firms should examine how work moves through the claims function, where responsibility sits and which activities add value.
Identifying unnecessary handoffs, duplicated tasks and opportunities for earlier intervention creates a clearer foundation for investment. Automated triage and watch lists can deliver incremental gains, but their impact is greater when they support a coherent service model.
Claims leaders should define the desired outcomes first, then determine how technology, people and process changes can deliver them. Success should be measured through client outcomes, decision quality and business retention, alongside processing efficiency.
Turn information into connected intelligence
Claims teams hold valuable information across files, emails, adjuster notes and other documents. Much of it has historically been difficult to analyse consistently or share with colleagues.
Large language models offer new ways to extract and interrogate this information. One migration example discussed at the conference reported a process running 270 times faster, illustrating the potential to reduce labour-intensive work.
However, accessing information is only the first step. Firms still need to establish its accuracy, provenance and relevance, supported by appropriate access controls and auditable processes.
For delegated authority business, the opportunity also depends on connecting information across MGAs, brokers, insurers and reinsurers. Extracted data becomes more useful when it can be shared consistently, with clear validation and traceability. Open APIs and integration with existing platforms can help reduce manual reconciliation and support gradual modernisation.
For specialty classes with relatively few, high-value claims, better access to existing knowledge may offer more immediate value than sophisticated predictive modelling. Identifying attritional claims trends across a portfolio can give underwriters earlier insight into changing performance.
Give human oversight a clear purpose
Complex London market claims require interpretation, commercial awareness and empathy. These capabilities remain essential when handling disputed coverage, multiple stakeholders or sensitive client circumstances.
Effective oversight needs more than someone approving an AI output. Reviewers need sufficient information, time and authority to challenge recommendations, with clear escalation routes when uncertainty arises.
Firms should also monitor whether reviews are meaningful, rather than becoming routine approvals. Automation can release capacity, allowing experienced professionals to focus on cases where their judgement makes the greatest difference.
Connect claims across the policy lifecycle
Claims expertise can contribute before a loss occurs. Wording reviews, scenario testing and discussions with risk managers can identify potential gaps early.
During the policy period, emerging claims patterns can inform risk management and underwriting appetite. At renewal, timely access to these insights can support decisions on pricing, coverage, limits and retention, alongside more informed conversations with clients and brokers.
Closing a claim should also trigger a feedback loop, ensuring lessons inform future decisions rather than remaining within the claims team.
Make adoption part of the investment
The gap between cultural preparedness and embedded change reinforces the need to involve claims handlers in design and equip leaders to support adoption.
Training should build critical thinking, confidence in interpreting outputs and an understanding of when to challenge them. Teams also need space to test tools, share feedback and improve their use.
The discussions at London Market Claims highlighted both the potential of AI and the organisational work needed to realise it. A consistent message emerged: lasting progress depends on redesigning processes, connecting reliable data and equipping people to use technology with confidence and sound judgement.
For the London market, the opportunity extends beyond faster claims handling. Bringing claims intelligence into underwriting, risk management and client conversations can strengthen decisions throughout the policy lifecycle. The next step is to turn that ambition into practical change, with clear ownership and measurable benefits for clients and the wider business.
To explore how better-connected delegated authority data can support your claims and underwriting decisions, speak to us today.
