Danish pension provider PFA is scaling up its use of artificial intelligence (AI) to identify customers at risk of long-term sickness absence, aiming to contact 5,000 individuals annually with early support interventions.

Since 2020, the provider has contacted nearly 5,000 customers flagged by its AI model, with around 1,500 going on to receive additional support following contact.

Among this group, PFA reported a 70 per cent reduction in the risk of progressing to long-term sickness absence, describing the results as “remarkable”.

The model analyses data points linked to customers’ use of health insurance to identify patterns associated with elevated risk, before cases are reviewed by healthcare professionals who decide whether to initiate contact.

The pension company stressed that the technology is used solely as a decision-support tool and is kept separate from pricing and other commercial activities.

PFA CEO, Ole Krogh, said the technology serves as a supplement to human judgement.
“AI highlights patterns, but it is always people who make the decision to call and offer help. That is crucial for us,” he said.

PFA said it hopes its approach can be adopted more widely, estimating that a national rollout could prevent up to 2,800 new cases of long-term sickness absence each year while delivering broader socioeconomic gains.

“It’s about helping people in time. We are only at the beginning of what AI can do for prevention and healthcare, and we hope our results can benefit even more people,” he added.

---