Q&A: Why Smart Insurers Are Betting Big on AI, And Getting It Right 

From wrangling messy data to navigating ethical minefields and regulatory twists, insurers making real progress are the ones getting the basics right. We caught up with Toby Fennemore, Solution Architect, Pro Global to find out how the smartest players are building momentum, and where the rest risk falling behind. 

How are insurers tackling data quality, regulation and ethical risks with AI? 

Toby: “AI is making headway in allowing us to transform the data into a common format from unstructured data, which is a significant problem in insurance, in order to fully leverage the capability we have to be realistic and map the as-is process and what we can do to streamline the need for data transformation . Insurers have made progress by focusing on data cleansing and structure before applying AI.  

But deep-learning models, Generative AI and LLMs add complexity – they’re not always transparent, and that’s a risk when you’re dealing with personal data. Ethical use, regulatory compliance, and model explainability are not nice-to-haves – they’re non-negotiables. The firms getting this right are those treating data quality and governance as strategic priorities, not IT problems. AI is also an environmental challenge, business need to make sure they understand the impact of their AI initiatives on the Environmental side of their Environmental, Social and Governance (ESG) policies “ 

Q2: What’s the smartest way to approach AI in 2025 and beyond? 

Toby: “Understand your processes before you digitise them. Delivering with AI is knowing what you want to solve. That means starting small, testing assumptions, and being crystal clear on how success will be measured. It also means investing in the boring stuff – clean data, consistent workflows, and good change management. Without that, AI is just smoke and mirrors. AI can’t solve bad processes, just do them faster. That said, the biggest gains will come from running two tracks in parallel: a disciplined, process-driven approach and a tech-minded, agile experimentation model. Alongside well-governed AI projects, insurers should also be running smaller, investigative AI pilots. Even when these fail, they offer huge value through insight – accelerating iteration and innovation. Controlled failure becomes one of your most powerful learning tools.” 

Q3: Is there a gap growing between AI adopters and AI leaders? 

Toby: “Absolutely. There’s a gap widening between those using AI and those who are embracing it at will. AI is becoming a central theme to Insurance Conferences, and those who are agile, experimenting with AI for both directed use case and R&D, will accelerate their business above those who aren’t. We need to be more tech-minded and embrace failure as a way to learn. Those companies who embrace AI and agile experimentation, will gain a wealth of data and knowledge of the best foot forward, while those still deliberating will be left behind.” 

Whether you’re AI-curious or already deep in the weeds of model governance and data pipelines, one thing’s clear: the insurers winning with AI aren’t waiting around for perfect conditions, they’re building the right foundations, failing fast, and learning faster. 

Find out more about Pro Global Digital Services.