FSCS chief: how AI will help us pay out claims quicker

Speaking to FT Adviser, Martyn Beauchamp said the lifeboat scheme was looking at using AI to help it scour through the tonnes of data and information it receives when investigating claims.

“There's a lot we can do with the kind of AI that's been around for a while now, such as machine learning and especially language processing,” Beauchamp said.

“If we can use that to pick up key words across thousands of pages of documents and PDFs and emails, and there is a way of doing that at scale, then this can help us shrink this amount of information to process and move things from the investigation stage as quickly as possible. 

“That's something we are considering and thinking about internally [at the FSCS].”

Data is a significant issue for the FSCS, as the amount of information its staff must process to get to the stage of paying out compensation has grown tremendously.

Beauchamp said the amount of evidence the FSCS gets on average has gone up 89 per cent over the past two years.

But it is not only the volume of evidence which is the problem; he said there were also issues and delays in collecting the information needed to carry out an investigation.

We want to move to a new advice model that enables us to have more control.Martyn Beauchamp, FSCS

“We usually get contacted by people because they are getting close to retirement”, Beauchamp explained; this means it could be years since the problematic advice was given or the problematic product was recommended.

In one example, Beauchamp said an individual had got in touch with the lifeboat scheme about nine years after he had received poor advice, which ended up being four or five years after the firm had failed.

He explained: “We have to go back in time, and by the time you deal with these, a lot of the people that know anything about the organisation have long gone and the IP probably isn't even there.

“Getting any information is so hard, so that's why we really want to improve on data collection. By bringing as much as we can in house, this will help us deliver better customer experience, productivity and a host of other benefits.”

Moving in-house

In a statement at the beginning of this year, Beauchamp set out how the FSCS would spend almost £9mn increasing its in-house expertise to deal with the increase in complex claims. 

This would mean the organisation could reduce the amount it spends on outsourcing in 2024/25. 

“We want to move to a target operating model for our advice claims,” Beauchamp explained.

“At the moment, two thirds of our operation is outsourced and our contracts for those end next year and so that will naturally require re-procurement.

"We want to move to a new advice model that enables us to have more control over the customer experience and more control over productivity.

“We're going to take the opportunity to flip that around and move from two-thirds outsourced to two thirds in-sourced. As you can imagine, that is the biggest change the organisation has been through in recent years.”

Customer experience

The FSCS has a special team to deal with vulnerable consumer concerns.

“We have 30 customer care champions in a unit and have people specially trained in dealing with vulnerable customers which is just  as important as dealing with complaints, Beauchamp said. 

“That's a good example of how we use expertise within the organisation to train new joiners. It's one thing dealing with complexity, but it's another thing dealing with people who really need your help. 

“You need to understand that some people are different and they have different ways of functioning and different ways of seeing the world and so our ability to respond to that is essential.”

He highlighted this as one area the FSCS wants to get better at, and has already improved its training programmes accordingly.

amy.austin@ft.com