Beyond Defensive Reporting: Machine Learning for Active Anti-Money Laundering Control in Insurance
A new paper explores the application of machine learning to actively combat money laundering within the insurance sector. Researchers trained gradient-boosted decision tree models using data from a Norwegian insurer to identify suspicious claims before payout. The study found that incorporating insurance fraud labels as an auxiliary training signal significantly improved the detection of money laundering cases, with the best model identifying nearly two-thirds of such cases within a small investigative subset. AI
IMPACT This research demonstrates a novel application of machine learning for fraud detection in the insurance industry, potentially improving risk management and regulatory compliance.