Business travelers on blast: Employees use AI chatbots to create fake expense receipts

Employees are increasingly using AI chatbots to create fake expense receipts for fraudulent reimbursements, exploiting easily accessible tools like ChatGPT to generate authentic-looking restaurant, hotel, and transportation bills. This emerging form of workplace fraud is becoming harder to detect as AI-generated receipts become more sophisticated, forcing some companies to revert to paper-based systems while others invest in new AI-powered detection tools.

The scope of the problem: A recent PYMNTS study found that 68% of organizations encountered at least one fraud attempt through their accounts payable services, including fake employee receipt submissions.

• The practice involves using free online chatbots to create bogus lunch receipts, hotel bills, and transportation expenses that employees submit for company reimbursement.
• Word is spreading rapidly on social media platforms about how authentic-looking these AI-generated forgeries appear and how difficult they are for employers to detect.

How easy it is: Creating convincing fake receipts requires minimal technical skill and can be accomplished in under a minute using free AI tools.

• One novice reporter managed to create a passable fake receipt in under a minute using basic prompts.
• More sophisticated results can be achieved by specifying restaurant locations and including partial credit card numbers to make the forgeries more convincing.
• Tech sector employee Deedy noted on X: “You can use (ChatGPT) 4o to generate fake receipts. There are too many real world verification flows that rely on ‘real images’ as proof. That era is over.”

Company responses: Some employers are reverting to traditional paper-based expense systems to combat digital fraud, while others are investing in AI-powered detection tools.

• “Lock it up, and get out,” said one Reddit user. “We’re going to a full paper based system again. The future is the past.”
• Tech companies including Expensify, SAP Concur, and AppZen are developing or already have tools specifically designed to spot AI-generated fake receipts.

Detection technology: New anti-fraud systems deploy multiple AI agents to identify telltale signs of fake receipts through various detection methods.

• These systems look for metadata fingerprints that AI bots leave on generated images, mathematical errors in calculations, and logical inconsistencies like incompatible items appearing on the same bill.
• AppZen cofounder and CTO Kunal Verna explained: “(O)ur Mastermind AI models work together, creating a platform of checks and balances. Where one model might miss a forgery, another catches it.”

What AI assistants are saying: Some AI platforms are beginning to flag suspicious receipts when asked to analyze them.

• Microsoft’s Copilot identified a fake Texas Roadhouse receipt, noting: “The receipt appears to be fake. The listed items and prices are unusual and do not match the typical offerings and prices at Texas Roadhouse… These items are not found on the official Texas Roadhouse menu.”
• This development suggests that AI detection capabilities are improving as the technology evolves.

Traditional fraudsters’ perspective: Some practitioners of pre-AI receipt forgery say they’ll stick with older methods like Photoshop and physical receipt printers.

• One Reddit user noted: “Receipt printer from AliExpress: $20. Photoshop license: $89 Unlimited expense receipts: Priceless.”
• However, these methods still cost employers money through fraudulent reimbursements, regardless of the creation method used.