Artificial Intelligence (AI) has successfully identified four out of five retirees who will retire on a disability pension, two years before they retire, the Finnish Centre for Pensions has revealed.

The pension body tested the machine learning algorithm on the centre's anonymous register data of 500,000 people, correctly predicting 78 per cent of future retirees who were set to retire on a disability pension in two years.

The tests were based on socioeconomic, earnings and benefit data held by the Finnish Centre for Pensions and were most successful at predicting disability pensions for those suffering from musculoskeletal diseases and respiratory issues.

Finnish Centre for Pensions mathematician, Jarno Varis, said: “Nearly four out of five is a very encouraging result. It is very likely that the accuracy would be improved by adding more social and health data to the algorithm.

“A low educational level, unemployment and being unmarried also contributed to the likelihood of retirement on a disability pension. The factors identified by AI correspond to the results of previous research.”

Furthermore, factors such as high age, repeated sickness benefits, rehabilitation allowances and reduced earned income were important indicators.

Finnish Centre for Pensions director, Mikko Kautto, said: “The extensive registers of the Finnish Centre for Pensions offer a good platform for using AI in the information industry. Once it gets easier to combine more extensive datasets, AI will offer improved tools to help identify risk groups and plan preventive methods.”

The research was conducted in partnership with Siili Solutions.

Share Story: