Briefing: AI must be used intelligently to augment people-based insurance sector

By Kelly Ogley

Reading businessman Elon Musk’s comments last November saying that artificial intelligence (AI) would eliminate all work made me feel incredibly uncomfortable and question how far AI will take us in the future. Is it a real threat to future roles in insurance, or are we naïve to think we will always need people?

The term AI is used a lot and we often associate it with ChatGPT and large language models (LLMs), but it covers many use cases based on machines that learn, reason, self-correct and predict.

Can machines really apply these skills successfully in all situations? There are many examples in insurance where they can and already are, including refining client contact strategies, settling claims, answering questions, pricing, extracting data and modelling. The list is endless.

AI, therefore, is certainly enhancing our industry and improving service and outcomes for our collective customers, which must be a good thing. So, how far do we think AI will go?

Does AI have the answers?

Insurance is about a promise and the fulfilment of that promise is built on trust. Yes, there’s a policy that outlines the detail, but fundamentally it’s all about trust.

One theory could be that there’s no better trust than that from AI because it will follow the rules to the letter and if the right conditions are validated, then the claim is settled – precisely and expeditiously without bias.

That said, flawed or biased data used to train AI can also lead to prejudiced outcomes and we need transparency in decisions.

As important as speed of claims resolution is, that is not all we are here to do. We are also responsible for listening to a client and understanding the verbal and physical cues that indicate they’re not sure about something or need more help.

Although voice AI that can pick up on emotions and change the direction of the risk capture or add in supplementary questions could lead us to believe that maybe AI can do it all after all.

But, to test that theory, we should question whether AI can really provide adequate understanding and support following a devastating incident.

AI can’t show empathy like a person can. It can be programmed to mimic or simulate emotions, but these are based on algorithms and pre-defined rules rather than genuine and authentic emotional understanding.

As for showcasing creativity and innovation, machines can learn, so they can innovate new products based on their experiences of losses, or forecasts of macroeconomic and geopolitical changes.

But what happens when the unexpected happens, such as a global pandemic? Does AI have the experience or empathy to know how to react?

Augmenting people

AI is and will continue to be a great addition to our industry, eliminating repetitive tasks, answering frequently asked questions, analysing vast amounts of data and improving efficiency and productivity. All of this should improve customer journeys and pricing by removing unnecessary costs and helping to ensure insurance is affordable for all.

So, let’s use AI intelligently and appropriately to improve and augment our people-based industry and not replace it.

I recently read about courier organisation DPD’s issues with its chatbot, where a customer frustrated with nonsensical responses managed to turn the chatbot against DPD, making it swear and write a poem about how bad the company’s service was.

If DPD had taken a co-pilot approach to its chatbot, getting the best out of people and AI, it would have had a better customer experience here and certainly less bad PR.

So, back to the question of whether AI will eliminate all jobs. I don’t think it will.

Insurance is a people-based industry, always has been and always will be.

No AI can arrange insurance for a Grade I listed property that holds weddings, has a hobby farm, has thatched buildings on its estate and is used as the client’s holiday home in the way an experienced broker and underwriter can.

No one can support a client at the point of claim like an empathetic, experienced insurance professional.

It’s crucial we don’t rest on our laurels and we embrace the amazing things AI can add to our industry, while also continuing to build great relationships, innovate together and enjoy ourselves along the way.