For the fourth time this year, Insurtech UK, facilitated by NashTech, led a review of the Product and Tech group's journey since July, including defining agentic AI, exploring its fit in insurance, and addressing challenges like data governance and legacy systems. The webinar highlighted the evolution from initial discussions to use cases and practical outputs.
What follows is a summary of the discussions during the 1-hour webinar and key takeaways for businesses considering agentic AI use cases in insurance.
Interested in joining the discussion? Get in touch.
The discussion was facilitated by NashTech’s, Chris Weston, with input from senior leaders from the likes of QBE Insurance, Oxbow Partners, Hastings Direct, MetLife and Capital Law, to name a few. The group shared how perceptions of agentic AI have shifted, with a focus on regulatory concerns, data quality, internal adoption, and the balance between innovation and risk aversion in the insurance sector.
Coco Co, Head of Finance Transformation from Hastings Direct, shared her learnings from a recent insurance conference this month, stating.
“Many have already adopted agentic AI internally because it improves efficiency and speeds up processes. However fully autonomous customer-facing deployments remain rare, largely due to regulatory uncertainty and trust concerns. The FCA has issued principles-based guidance focused on customer outcomes rather than prescriptive rules, which leaves firms cautious about scaling agentic AI for direct customer use.”
Coco reiterated that the focus remains on using AI as a tool to support human workers rather than replace them, with empathy and customer trust seen as critical.
“Insurance remains a human business. Speed brings down cost, but empathy keeps your customer.”
It was continually noted that the insurance industry is traditionally risk-averse, leading to slow investment in new technologies. Many organisations prefer a 'fast follower' or 'watch and learn' strategy, waiting for others to demonstrate success before committing significant resources.
Simon Liste, Apollo Underwriting emphasised the need for cultural adaptation, governance frameworks, and clear accountability as agentic AI moves from augmenting existing processes to automating tasks. The transition requires careful management of both technical and human factors.
Simon Liste, Head of Core Technology at Apollo Underwriting added, “Major technology providers do not believe they should be accountable for setting governance and controls for AI, instead relying on early adopters to learn from mistakes and help shape future regulation. This creates a reluctance among organisations to be early adopters, as no one wants to be the first to fail and become a regulatory lesson.”
Charlotte Gregory, Partner at Capital Law, added, “I sympathise with the nervousness around outcomes focussed regulation. My personal view is that firms won't really get detailed views from the FCA on what 'good' looks like (beyond existing guidance such as the FCA’s AI update) until it has some 'lessons learned' to comment on or regulatory action is taken. No one wants to be that use case. Starting with a simpler product or a contained process makes a lot of sense to me. As many have said using the right tool for the matter at hand is key”.
Kai Wing Shiu, an Independent Consultant said, “Organisations should invest in learning and experimentation even without immediate payback, this will position them to respond effectively when the technology matures, and market conditions shift”.
Charlotte Gregory added, “One thing to keep a weather eye on is the UK government open call for evidence on its proposed AI Growth Lab. Businesses may consider responding to see what insights can be gained, and what other sectors are actually doing (or not doing) when implementing AI tools into their businesses”.
Overall, the Working Group’s journey demonstrates the importance of collaboration, stakeholder engagement, and ongoing learning. By embracing a cautious yet proactive approach, the insurance industry can harness the benefits of agentic AI while safeguarding customer interests and regulatory compliance.
Insurtech and NashTech are looking forward to meeting next month with the community to share our latest findings with the Agentic AI in Insurance Whitepaper.
This article and any associated discussions were conducted under the Chatham House Rule. No comments or insights are attributed or reported unless expressly permitted by the participants.