Machine learning presents a solid opportunity for the industry and regulators to fight back against fraudsters. By automatically combining hundreds of variables from different datasets, including patient/prescriber history, interaction graphs, prescription characteristics, and other contextual data, it became possible to target actual fraud cases.