Two studies released in mid-2026 show AI-generated fake expense receipts have rapidly become the dominant form of expense fraud. AppZen platform data reported in Accounting Today shows the share of flagged fraudulent receipts that were AI-generated went from 0% in March 2025 to 70.8% by mid-May 2026 — based on 1,471 AI-generated fakes from 745 employees at 174 companies totaling $148,143 in fabricated claims. A separate Emburse survey of 2,000 U.S. and U.K. workers (conducted by Atomik Research) found 40% of U.S. respondents admitted to generating a fake receipt with AI, with 40% of those using company-funded AI tools to do so.
Two independent data sources published in mid-2026 both point to the same shift: AI image generators have become the primary tool for expense receipt fraud. Accounting Today, citing AppZen platform data, reports that the share of flagged fraudulent receipts identified as AI-generated went from 0% in March 2025 to 70.8% by mid-May 2026 — a figure based on 1,471 AI-generated fake receipts submitted by 745 employees at 174 companies, claiming $148,143 in fabricated reimbursements. This figure describes the composition of detected fraud, not the overall fraud rate. A separate Emburse survey of 2,000 workers conducted by Atomik Research (May 5-8, 2026) found 40% of U.S. respondents and 29% of U.K. respondents admitted to generating a fake receipt using AI. Globally, 40% of those who did so used company-funded AI tools, and 9% built their own tools for the purpose.
According to Accounting Today's reporting on AppZen data, the crossover from template-based fraud to AI-generated receipts happened in April 2026, when AI-generated fakes first exceeded templates. Previously, templates from specialized websites accounted for 95-100% of flagged fakes. The Emburse survey breaks down use types: 19% of respondents have used AI to completely fabricate a purchase, 15% to inflate the value of a real expense, and 6% to replace a genuinely lost receipt. AI-generated fakes average roughly $100 per receipt and a median of $32 — about half the average $182 for older template-based fakes — consistent with a high-volume, low-dollar strategy designed to stay below auto-approval thresholds.
Traditional image-based receipt verification relies on visual artifacts that older template tools left behind. Modern AI image generators produce receipts with realistic textures, plausible itemization, and accurate layouts without specialized skills, making purely visual inspection less reliable. Accounting Today notes that 32% of finance professionals in the U.S. and U.K. reported they would not be able to recognize an AI-generated fake receipt. "AI makes it easier than ever to fabricate receipts and bypass traditional controls, so organizations need a more proactive, AI-centric approach to managing spend," said Michele Shepard, chief revenue officer of Emburse.
Platform data and survey data converge on a clear structural shift: AI image generators have displaced template-based tools as the primary method of expense receipt fraud, with the share of flagged fakes reaching 70.8% in under 18 months. Notable for practitioners in compliance, finance operations, and fraud detection tooling. Not a frontier AI research story, but a concrete deployment-level impact story with well-sourced numbers.
A 5-minute Tuesday brief on AI & data science. Curated, no fluff.
No spam. Privacy.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
News on Let's Data Science is compiled from multiple public sources with editorial oversight. See our Editorial Standards and Corrections Policy.
Leave a Reply