Denmark’s largest pension fund, PFA, said its artificial intelligence (AI) model used to identify scheme members at risk of long-term illness – and the action it has taken based on the data – have led to a 70% lowering of that risk.

Morten Winther Hansen, chief operating officer of the DKK736bn (€98.7bn) pension fund, said: “It is a very significant effect, which has taken us completely by surprise.”

PFA had been working with AI for several years, he said, and had put together “a team of some of the country’s smartest minds”.

“Everything from quantum physicists to people with data science PhDs have now created an enormously useful tool that we can see works,” he noted.

PFA said the system worked by the AI model first notifying its healthcare staff if a customer showed signs of risk of becoming long-term ill. When this happened, PFA would contact the customer to ask how things were going and whether they needed help, it said.

Over the past two years, it said 1,700 customers had been called in this way, which had significantly reduced their risk of long-term illness.

Winther Hansen said society was facing a health crisis, with more and more people becoming chronically ill.

“It is a huge challenge that PFA is helping solve by preventing and helping customers with early and proactive action.

“Here, our preventive use of artificial intelligence plays a decisive role, and this is where we as an industry have something special to contribute,” he said.

The pension fund said its AI model only used standard data customers had provided, and that the model used a total of 29 data points and their interactions to calculate a customer’s individual risk of long-term disease.

Read the digital edition of IPE’s latest magazine