Expense report fraud accounts for 14% of occupational fraud cases, with the median loss estimated at $33,000 (source: Association of Certified Fraud Examiners, Report to the Nations 2024). With the rapid advances in artificial intelligence, CFOs and accountants fear a surge in attempted fraud cases.
In the past, employees needed to be proficient in image-editing software to discreetly alter the amounts on a gas receipt or a restaurant bill, whereas today all it takes to generate a fake receipt is a simple prompt to any AI tool.
What are accountants and CFOs afraid of? That fraudsters will take advantage of the situation to artificially inflate their expenses and obtain an improper reimbursement from the company. Fake receipts are sometimes so convincing that even experienced auditors have a hard time spotting them.
AI-enabled expense report fraud typically involves: falsifying information on receipts (amount, date, fuel volume, etc.) or creating entirely fake expense receipts.
How does expense report fraud work with AI?
Generative AI tools can be used to modify existing documents, or even to create fake receipts from scratch, in ways that are sometimes completely impossible to detect.
“At first, the most common cases of fraud we encountered were changes to the total amount. Today, even VAT rates or specific details, such as fuel quantities, are being altered”, explains Bertrand Muth, General Manager of International Affairs at N2F.
“We’re also seeing an increase in IBAN-related fraud: some scammers change the account number on invoices to redirect payments to their own accounts,” he notes.
According to a study conducted by SAP Concur on business travel, more than two-thirds of finance executives and travel managers suspect that AI-driven fraud is taking place within their organizations, and one in ten is “certain” that such fraud exists.
According to the study, this phenomenon is part of a broader trend of increased expense fraud since the pandemic.
AI-powered fraud: a risk you need to be aware of
Back in 2021, another study conducted by SAP Concur revealed that employees considered it “acceptable” to claim an average of €114 in unsubstantiated expenses. The Association of Certified Fraud Examiners study, meanwhile, highlights that “schemes committed by an owner/executive have a velocity nearly three times that of schemes committed by employees and manager-level individuals.”
For finance teams, this issue is becoming even more critical as AI becomes more widespread and increasingly difficult to detect. As a result, organizations that still rely on manual expense report verification and validation processes are even more vulnerable.
“If a receipt or supporting document looks genuine at first glance, it can easily slip through the net,” warns Bertrand Muth.
AI vs. AI: technology for fraud detection
As you can see, AI has made it easier to forge receipts and invoices. It has also made the process more sophisticated. But this very same technology can actually be used to combat fraud.
Solutions such as N2F run automated controls, using AI and machine learning to identify suspicious documents. These checks go beyond financial amounts, they also analyze VAT rates, fuel quantities on gas receipts, and other details.
For IBAN fraud, the system compares the bank account number with previous invoices to flag any anomalies. Human vigilance nonetheless remains the first line of defense.
At N2F, we recommend that you call your supplier to verify that their bank details are correct before making your first payment. This approach is known as “human-in-the-loop”.
Fraud with fake receipts: what does AI detect?
Solutions like N2F, which use AI to detect fraud attempts or irregularities, will alert you in the following cases, for example:
- Unusual structure or layout
- Inconsistent logos, fonts, or typography
- Artifacts in generated images
- Illogical amounts
- Receipts from countries that do not match business travel data
- Expenses incurred on a non-working day
- Duplicate expenses (including those incurred by another employee).
Going forward, systems based on autonomous agents could even assess whether receipts meet tax requirements. AI will become a strategic ally in the fight against expense fraud.
A game of cat and mouse
For Bertrand Muth, detecting documents forged with AI is an ongoing battle:
“It’s a real game of cat and mouse, similar to the battle between viruses and antivirus programs”, he explains.
Fraudsters are constantly coming up with new ways to circumvent controls, meaning that companies, must continually adapt their fraud detection methods
As Bertrand Muth explains, the best way to prevent fraud is to centralize all expenses using a corporate credit card, with receipts automatically linked to bank transactions.
We also recommend a multi-layered approach that combines technology, human oversight, and team awareness.
Best practices for combating expense report fraud
Here is our advice to effectively combat expense report fraud:
- Always use corporate credit cards for business-related expenses
- Link receipts and card payment records to bank transactions for complete reconciliation
- Manually verify the details of any new supplier before paying the invoice
- Combine AI-powered detection tools with human oversight
- Train teams on the risks and consequences of fraud.
💡See also: How to choose the right App to save time and simplify Expense Claims
N2F: AI for smarter and more efficient expense management
At N2F, we do not restrict AI to fraud detection. Thanks to the technology developed by N2F, managing expense reports is faster, more reliable, and more transparent.
A reliability score is assigned to each expense–anomalies and irregularities are flagged automatically. This means that the accounting teams can focus on the cases that really require their attention.
As you can see, fraud using AI-generated fake receipts is not a distant threat. It is a reality that is driving companies to modernize their expense management and strengthen their internal controls, significantly increasing their chances of detecting fraud in time.
What are the advantages of AI for managing expenses?
When used in business expense management tools, AI will help teams:
- Immediately identify suspicious expenses
- Detect frauds and inconsistencies
- Correct potential input errors
- Reduce expense report processing time and speed up reimbursement.
Want to learn more about how digitization and automation can help you control expenses and prevent fraud? We’ve put together a guide on how to choose an expense report solution!
