New technologies, including artificial intelligence and machine learning, increasingly are being applied to the old problem of occupational fraud. But in most circumstances, common accounting tools — “variance analysis” and “contribution margin” — remain effective in uncovering possible evidence of theft. Gaps and absences After your organization finalizes its annual budget, you may perform a variance analysis, reviewing differences between actual and budgeted performance. If, for example, actual wages significantly exceed budgeted wages, the difference could be due to such factors as wage increases, productivity declines or greater downtime. But it could also signal phantom employees on the payroll. Fraud experts pay particular attention to variances related to inventory and purchase pricing. Supply-related variances could indicate the existence of kickbacks. Or they might suggest fictitious vendors —...