QBE Survey: 80% of Large U.S. Companies Hit by Employee Fraud as AI Threat Grows
Wednesday, March 4th, 2026 Fraud Insurance Industry Litigation Risk Management TechnologyA new survey from QBE Insurance finds that 80% of risk managers at large U.S. companies experienced employee theft, fraud, or embezzlement in the past year. The poll of 200 risk managers at firms with at least $500 million in annual revenue highlights how traditional internal schemes such as billing fraud are intersecting with emerging AI-driven threats.
Billing fraud was the most frequently reported scheme over a three-year period, cited by 36% of respondents. Payment and check fraud followed at 23%, payroll fraud at 19%, and cash or non-cash theft at 23%. Billing fraud remains difficult to detect because it moves through the same accounts payable systems used for legitimate vendor payments. When invoices involve services rather than physical goods, the absence of inventory controls increases vulnerability. For claims professionals, this reinforces the need to closely examine documentation trails, vendor histories, and approval workflows when handling commercial crime losses.
Loss severity data from the Association of Certified Fraud Examiners underscores the financial stakes. Its 2024 Report to the Nations placed median losses at $145,000 per case and average losses at $1.7 million. The group estimates organizations lose 5% of annual revenue to fraud. For a $500 million company, that equates to $25 million in potential exposure. Senior leadership involvement drives larger losses, with median losses of $500,000 when owners or executives are implicated compared to $60,000 for staff-level employees. QBE’s findings align with that trend, reporting that more than three-quarters of recent incidents involved managers or higher-level personnel.
AI-related risk is now a central concern. Ninety-four percent of surveyed risk managers expressed worry about employees using AI tools to commit workplace crimes. Nearly half of organizations already deploy AI-based detection systems, and half of the remainder plan to implement such tools within 12 months. Deepfake-enabled fraud, including AI-generated voice and video impersonation used to authorize transactions, has heightened scrutiny of internal verification controls.
For adjusters, these developments affect both underwriting intent and claims handling. Crime policies are being tested by impersonation losses, social engineering claims, and disputes over computer fraud triggers. The survey found that 87% of companies carry crime insurance and nearly two-thirds plan to increase limits, reflecting heightened awareness of AI-powered fraud and economic pressure. Clear documentation of control environments, separation of duties, and authentication protocols will be critical in evaluating coverage applicability and potential sublimits.
The broader cost of fraud remains substantial. The Coalition Against Insurance Fraud estimates fraudulent activity costs Americans up to $308.6 billion annually. For the insurance industry, the data points to sustained demand for crime coverage, increased litigation risk over policy wording, and continued investment in fraud detection technologies that adjusters will encounter during investigations.
