Though the claims landscape may have changed almost beyond recognition since James Nicholson (pictured) joined Zurich Insurance as part of its graduate program 27 years ago, its premise of supporting local communities and wider society remains the same.
Looking back on his career to date, Nicholson – who stepped into his role as chief claims officer for Zurich in the UK one year ago – highlighted the lasting appeal of working in the claims sector and the genuinely inspirational people he has worked with.
With inflationary pressures and ongoing supply chain challenges still pressing issues for the market, Nicholson emphasised the role that advanced digital capabilities – and AI as a crucial aspect of that – will play in 2024. The question is where to place your chips and getting the right answer will require a “fail faster” culture and approach where new ideas are embraced, implemented and assessed rapidly, he said.
“It’s about being clear about what your priorities are and what success looks like to you,” he said. “I was on a call with EY recently who were talking about how 97% of banks have AI labs already in place and how the insurance market has been slow to follow that. One of the areas they’ve mapped out is that the three stages of AI are: advise, assist and coach. And we’ve played very much into that advise space and it’s the assist space we’re progressing into.”
An easy trap for carriers to fall into is trying to jump in at too high a level when it comes to AI, without getting the fundamentals right, he said, which is why having a clear vision is so important. For Zurich UK, that’s about being really clear that AI is not here to replace what its customer delivery proposition looks like, but rather to enhance that proposition.
AI has the capacity to disrupt and enhance the claims process in several ways, he said, and that’s not always to the benefit of customers or insurers as it has the potential to be used for fraudulent activities such as creating deep fakes or manipulating data. However, the solutions for these risks are also to be found in AI tools which can be calibrated to catch inconsistencies and anomalies that might go undetected by human eyes.
For Nicholson and his team, the objective is to find where AI tools can be deployed to maximum advantage, particularly in the context of enhancing the customer experience by facilitating chat facilities and speech analytics. If used correctly, AI has the capacity to quickly identify vulnerable customers and map out ways to support them in getting their claims on track to a faster resolution.
Additionally, he said, it can optimise operational transactional processes to improve process efficiency, and his team are working with Keoghs on utilising AI for quantifying injuries from medical reports. This work is already giving 20-to-25 minutes back to claims handlers, by automating the tasks that don’t require their unique skill sets. Report reading accuracy has already reached 93% for muscular injury recognition and work has already started on developing that further into some other casualty areas.
“It’s clear that the process efficiency opportunities around AI are really strong,” he said. “But we’re managing our progress very carefully so that our employees understand that we’re not taking away jobs from people and giving them to robots. It’s the non-value-add tasks where we can put AI into action but we don’t have any end-to-end AI solutions - we will always have a human in the loop somewhere along the claim.”
AI presents so many opportunities for the insurance market, he said, but successfully mitigating AI risks and harnessing these opportunities comes back to being very deliberate about what your priorities are. A clear vision and a clearly defined ethical framework for utilising AI is key, and with that as your guiding principle, you have a strong foundation on which to explore some of the opportunities available for the good of customers, as well as employees.
“The insurance market, including ourselves, is facing a crucial question – do you want to be a leader or a follower in this?” he asked.
“We recognise the need for a high level of transparency and honesty about how different digital tools are working or not working, and to use those findings to continually propel us to the next level. And I think the insurance industry as a whole needs to be a little bit more culturally aware of the necessity of that fail-fast approach.”
