Artificial intelligence (AI) and its capacity to transform the insurance industry was on the mind of the market during Insurtech Insights’ 2025 European conference. In a panel discussion on ‘Revolutionising Insurance Distribution’, insurance specialists shared their perspectives on two pressing questions – how should we be thinking about the relationship between human beings and AI, and where does the industry go in the next five years?
Offering his insight, Tim Hardcastle, INSTANDA CEO and co-founder, noted that he has seen multiple different technology ‘waves’ throughout his career and that while the AI ‘wave’ is similar in many respects, there are some crucial differences. Take, for example, he said, the dot-com boom and, later the shift from on-prem to Cloud. Each of these advancements went through cycles, and it takes time for those cycles to wash through. That timing is largely impacted by two key elements – the physics and the economics.
What he finds fascinating about the AI evolution is how rapidly those factors are developing. “Until a few months ago it was thought that the investment needed for AI evolution is massive, but now we’re seeing much cheaper models being produced, even by companies like Microsoft,” he said. “Then, separately, on the penetration or rate of physics, there’s no barriers in terms of its ability to reach people in industry and society.”
As to what the next five years is likely to bring, he advised on the need to keep watching what the big tech companies are doing and the advancements they’re making to their own models. Open AI is actively looking at its own model and how it interprets data and draws conclusions not just for the ease of users, he said, but also because they want to understand the reasoning process and its rate of evolution for themselves.
“One thing we can say of this rapid pace of evolution is that the human elements of interaction with this technology will migrate to the more sophisticated and nuanced end of decision-making,” he said. “[That’s around] helping to frame some morality in the way conclusions are reached and making sure that discrimination is limited. The role of the human is where I think we’ll see a hybrid model evolving over the future.”
From an insurance context, he said, it’s important to register the potential for AI to bridge the protection gap in areas where there’s limited penetration, and to bolster claims efficiency. “Aren’t we really fortunate to be looking out over the horizon of the next five years and thinking about all the wonderful possibilities that are in front of us? Because the technology is certainly not going to be a limiting factor.”
Melissa Wong, group chief product officer at bolttech, emphasised that AI has the potential for significant transformation in insurance given the “unnatural” nature of the insurance proposition. Insurance is something customers buy on the chance that they’ll need it, she said, and claims is the moment of truth for that promise. “So, when you think about how AI can impact this whole industry, it’s really even before insurance – it’s [about] those preventative, predictive aspects.”
Those aspects can be supported and enabled by AI, she said. There’s also significant opportunity for the industry to start exploring how to use AI in more humanistic ways to make a tangible difference to our societies. The impact of doing so is not limited to creating better predictive and proactive capabilities but will also support the other end of the insurance process. “So, when the bad thing happens, how do you assist and remediate and support those human beings who are actually experiencing that. With that, I think there's a fascinating ability for AI to actually provide scaled service in areas and geographies that we are not able to access.”
Adding his thoughts, Balazs Gati, chief product and technology officer at the digital brokerage Clark, emphasised the need to address current challenge around the deployment of AI within organisations. When you have AI replacing low-skill tasks in the fields of operations, underwriting and claims, it becomes very difficult to coach people on the job, he said. “There’s always this vision of AI agents doing the simple tasks, with an expert team leader overseeing and supervising these agents. Currently that equilibrium is there but over time, it [represents] a challenge.”
The challenge that Gati sees is how to bring in the right pipeline of junior talent – and provide the right support and training that will allow them to grow into supervisory roles – if they don’t have access to the experience required to develop those technical and managerial skills.
Echoing the need for a considered approach to the opportunities represented by AI, Tom Thornberry, group chief claims officer for Zurich Insurance Group, underscored the need for these tools to be utilised “responsibly, for the right reasons.” What customers essentially want from insurance is speed of certainty and an empathetic approach. But when you look at the nuances of that, he said, it varies from product line to product line, and becomes more complex every layer you go down. “We’ve got a responsibility to utilise this effectively for the right reasons. So, where we’re putting it in from a claims perspective, it needs to bring that speed of certainty. It needs to make sure we’re driving the right behaviours so that we’re delivering the best possible service and claims to our customers.
“But it also needs to be done in a way that allows for our people to have a high quality of job satisfaction, eliminating repetitive tasks but also taking into account [the need] not to create a gap… We need to make sure that we are deploying technology in the right places for the right reasons and to do so responsibly. I think over the next five years, that question of responsibility is going to be critical, and I think that's where we really need to make sure, as an industry, we are aligned.”
