AI revealed more loan fund cheating

Photo: Thomas Brun / NTB scanpix

Artificial intelligence revealed more loan fund cheating

The Norwegian student loan fund (Lånekassen) has for the first time used artificial intelligence (AI) to select account controls. It provided twice as many revelations of cheating.

The book check is done among students who state that they live away from home, since only students who do not live with their parents can get the loan back from grants.

Lånekassen’s book check in 2018 included 25,000 students. 15,000 were picked out by artificial intelligence (machine learning), while 10,000 were picked out by random selection (control group). The results showed that artificial intelligence made it easier to find those who are not really living away from home, said the Loan Fund.

‘’5.5% in the control group and 11.6% in the machine learning group did not pass the book check. It shows that the machine learning selection has been twice as effective as the random selection in finding students who have given us incorrect information about housing,’’ said CEO, Nina Schanke Funnemark, of Lånekassen.

She pointed out that this method based on knowledge from previous controls means that the Loan Fund can not control those who, with very high probability, are real ‘away dwellers’.

‘’Thus, fewer students have to take the trouble to submit documentation, which also means less case processing with us,’’ said the director.

Soon it’s time for new book controls. This time too, artificial intelligence will be used.

© NTB Scanpix / #Norway Today
RSS Feed