Sponsored content: Mohit Manchanda, SVP and EMEA business head for insurance, healthcare and life sciences at EXL, discusses how AI is moving from idea to implementation in the insurance sector
How is the UK insurance industry using artificial intelligence (AI) to accelerate growth and improve productivity? How does its success compare to other industries? We explore these questions and more in EXL’s second report on the state of AI.
In 2025, we surveyed over 190 companies, including 33 insurance firms, on how they’re using AI.
Insurance companies in the UK reported a 21% reduction in operating costs with AI – the highest amongst all industries surveyed – and 18% revenue growth, which is same as in banking and utilities.
Embedded into workflows
The insurance industry is deploying AI across front and back office operations. AI technology, such as generative AI, and machine learning have been the most deployed capabilities in customer care (58%) and finance and accounting (55%).
Additionally, 40% of companies are using AI in claims and risk management.
We see that AI is not replacing human agents – it is augmenting them by processing the tasks, or by analysing high-performing agents and making them successful.
Underwriting taps AI
Historically, underwriting has been a document heavy, manual process. Today, AI models can extract hundreds of data fields from submission documents, reducing many hours of manual processing.
This capability does more than just save time. It helps insurance agents carry out risk assessments with greater accuracy, classify risks into the right business class or SIC codes and expedites communication with brokers.
By automating the ingestion of complex data, insurers can free their skilled professionals to focus on the nuanced decision making that drives profitability.
Similarly, customer service processes are applying AI to enhance customer experience and agent productivity.
AI identifies strong leads for brokers to focus on while also supporting all agents with nudges that convert low probability leads to high probability.
Additionally, AI identifies high-performing agents, uses their call transcript to develop nudges that can assist agents needing improvement. These insights are then developed into behavioural nudges that assist other agents who may require improvement, raising the standard of service across the board and adjusting strategies.
For a real-life example one motor insurer focused on streamlining email communications between partners and customers within the claims function. Their system suffered from inefficiencies due to numerous mailboxes, missing information in emails, multiple channels and inconsistent manual follow-ups.
Using agentic AI, the insurer automated follow ups by augmenting missing information in real-time, integrated with underlying system and triaging emails to the correct mailboxes for further downstream actions.
Insurers are clearly committed to developing their AI capabilities. Results show, 79% indicate that achieving real scale in this area is a priority.
The path forward
While AI is helping brokers, insurers, and MGAs work smarter, not harder, there is still more to accomplish with agentic AI – and trends are promising.
While transitioning from pilots to production is challenging, the data suggests that UK insurers are beginning to bridge that gap by focusing on high-impact areas and building workflows and platforms that are natively agentic.
- To learn more about these trends and view the full data, click here
