By Glenn Marshall
I’ve been thinking about this for a while now, and the more I hear and learn about AI, the more I think we need to stop kidding ourselves and have a proper conversation about what it really means for our market. There’s a lot of chat about AI at the moment, mostly around cost savings, speed, efficiency, accurate data and taking some of the grind out of the day job.
Fair enough. Some of that is true. I can absolutely see where it could help. It can sort through submissions, pull out the key points, flag what’s missing and take some of the repetitive work off people’s desks. No one’s going to miss the mind-numbing stuff.
But that’s not really the point for me. It’s all about the long game. What does this mean for underwriting as a profession? And what does it mean for the younger generation coming into specialist insurance? This job isn’t something you learn by pressing a button or reading a guide. I can still remember sitting with John Murphy, asking stupid questions and learning from people who really knew their trade. I was lucky to be surrounded by some proper mentors early on, and that’s how you learn this job properly. You listen, you watch, you ask, and over time you start to get a feel for what matters and what doesn’t.
You realise that what looks fine on paper isn’t always fine in the real world. That only comes with time, exposure and experience. It doesn’t come from a machine giving you a tidy summary… which is often a load of rubbish. Try asking AI, “Should I drive to the car wash that is only 100 metres away to get my car cleaned or should I walk?” Both ChatGPT and Claude told me to walk!!
Now, I’m not completely anti-AI. Used properly, I’m sure it could be a useful tool. It could help underwriters get through the noise quicker and spend more time on the bits that actually need judgement. It could help smaller businesses work smarter and free people up to focus on brokers, trading and the risks that need proper thought.
That all makes sense. What does concern me is, entry level, early career and back office administrative roles are likely to be the first ones in the firing line. Where do these employees go after this happens? This market has always given conscientious people an opportunity and a way in. I don’t know many in insurance who turned up with a shiny CV and some grand plan. Plenty of us came into this business due to friends and family already being involved. Started from the ground up and built careers through experience, support and hard graft. That’s one of the best things about this profession; you can start at entry level and really grow. I do not see how AI offers this path.
But if those starting points begin to disappear, or get watered down too much, then what? Where do the next good underwriters come from? Where do they get their grounding? You can’t keep banging on about talent and the future of the market while getting rid of the very jobs that help people get started. That’s what we need to be careful of. Because if we’re not, we’ll end up with a market that looks the part on the surface but underneath has no depth. Faster processes, less real experience coming through, and more people relying on what a system tells them instead of having the confidence experience and knowledge to challenge it. In specialist insurance, that would be a big mistake, huge!
However good technology gets, this is still a people business. It still comes down to judgement, instinct, experience and having the bottle to make a call. A lot of this game sits in the grey. That’s why good underwriters and a good team matter.
So no, AI should not replace underwriting decisions. Help it? Yes. Support it? Definitely. Replace it? Hopefully not!
The real danger is if it strips away too much of the learning and early experience people need, because that’s when the market will start to feel it. Maybe not tomorrow. Maybe not next year. But a few years down the line, when people start asking where the next generation of proper underwriters have gone.
If I’m honest, that’s only scratching the surface because what worries me about AI goes way well beyond underwriting and the industry as a whole, it’s about what happens when we hand over too much, trust and only realise too late what it’s doing to work, to the economy, to people and to the world around us.