AI-Generated Fake Receipts Now Trigger 70% of Fraud Flags

TLDW: AI receipt fraud has surged from zero to 70% of detected fraud in 14 months, driven by employees using company AI tools to create fake expenses ("revenge spending") amid job displacement fears. Key points: - AI receipt generation has become trivial—hosts demonstrated creating convincing fake Walmart receipts in early 2025 with readily available tools - "Revenge spending" emerged as a new fraud motivation where employees exploit AI to generate fake expense reimbursements before being displaced - The problem exploded from nonexistent in February 2025 to accounting's dominant fraud type by June 2026, now representing 70% of active fraud flags - Controllers and accountants are described as facing a "distressing" emerging threat requiring urgent attention and new detection protocols - The rapid adoption of company-provided AI tools by employees has inadvertently become a vector for internal fraud rather than a compliance asset

Continue reading

Get daily agentic AI accounting news in your inbox
Read original article →
View Transcript

Hello and welcome back to the accounting podcast, your weekly roundup of news in the profession. I'm Blake Oliver

and I'm David Liry.

And David, our lead story today is all about how AI receipt fraud is soaring. People are using AI to create fake receipts. We've got a couple of surveys on this really distressing something controllers and accountants need to be paying attention to. We will get into that right after we thank our sponsors.

Our sponsors this week, we have cloud accountant staffing, valuebuilder systems, and onpay.

Are you tired of the endless search for qualified accounting talent? You're not alone. Growing accounting firms are struggling to find available and affordable team members when they need the most. Cloud Accountant Staffing has the solution with their revolutionary candidate portal. Unlike traditional staffing agencies that waste your time with sales calls, paperwork, and deposits, Cloud Accountant Staffing Candidate Portal gives you instant access to highly vetted, qualified accounting professionals. No waiting, no hassle, just top talent right now. What makes this different? Speed and simplicity. While other firms make you wait weeks or months, with cloud accountant staffing, you could interview someone as soon as tomorrow. Their boutique support ensures you're getting quality talent that's both available and affordable. Exactly what growing firms need. The candidate portal puts you in control. Browse live candidates, make selections on your timeline, and build your offshore team without the traditional headaches. To find, review, and book interviews with potential team members, all in less than 10 minutes, head over to accountingpodcast.promo/cas. That's accountingpodcast.promoc promo slc. And now David, I'm going to let you start with this story. We both saw this AI fake receipts. What are the numbers? What are we What are we talking about? How bad is this?

Yeah. So, before we jump to the numbers, a new term bubbled up in one of these articles that caught my eye and it's the term of revenge spending.

Revenge spending.

So, revenge spending, that is when employees think AI is going to replace their jobs. So they use the company provided AI tools to create fake AI generated receipts and expense reports

and then they put these in and get reimbursed for fake expenses. It's uh sort of like the um modern equivalent of like stealing office supplies. Sounds like

pens. They're taking pens. Yes.

Yeah.

Well, what is this like? Yeah. How bad is it?

It didn't exist. Think about it. Um March of 2025, I think sometime in February 2025. I think even on the show we live created like a fake Walmart receipt.

Yes.

With AI.

Yeah.

And it was good enough. It wasn't great.

But so it went from zero people doing this to 70% of fraud flags now are popping up. So it just in 14 a little 14 months it's exploded as a big issue for um accounting departments. Now the numbers are even crazy based on these surveys. So these were two surveys from uh one is the APN expense management app A appzen and then another one is Embers who's also a travel expense solution provider. So they have their own data. Um so one of the crazy ones is 40% of US workers admit using AI to generate fake receipts.

40% that's

and 40% of those said they use their employer employer funded AI tools to do it.

Wow.

You're not even like trying to take it offline. you're just doing it at your desk using the computer and the tools the uh company provided you.

Now, these are not like enormous amounts. According to that AppZen survey, AI generated fake receipts averaged about $100 each with a median around $32. And the older template based fake receipts that people used to make averaged $182 each. So, it looks like we're getting more fraud but smaller dollar amounts. And the what the reason they're doing the smaller dollar amounts is because all these apps ramp etc have all these tools to auto approve stuff and you have AI routing things around. So if you can make your receipts not hit that threshold it'll just automatically get pushed through the system.

But now what about these tools that you know RAMP is adding in other other spend platforms are adding in. I know Ramp is doing it that will detect fake AI receipts because you can detect them

and that's what's happening, right? It's being, you know, that's up 70% the detection of these.

Oh, that's the stat you mention.

70%.

I almost feel like it's it's similar to um spam, right? There's AI and technology makes it easier than ever for people to send you millions of spam messages. But then on your side, you're using all these AI tools to detect the spam messages and move them to your trash. It it's it's kind of like that. The AI on one hand is making it very easy for employees to cheat and create fake receipts, but the at the other hand is hopefully helping you catch them. So there's no benefit to using AI. I guess it's a wash at the end of the day, right?

And it's not that employees are necessarily creating totally fake expenses. That is happening, but there's a variety of different kinds. And the Embers survey detailed this. So found a few different types of misuse. 19% of their respondents who admitted to creating fake expenses said they had used AI to invent a purchase entirely. So about one in five. 15% said they used it to inflate the value of a real purchase. So real purchase, but they just increased the dollar amount. 6% said they used it to recreate a loss receipt for a legitimate purpose.

I was going to do that this week. Honest to God, David, I was going to do this.

Don't do that to me. And and then um like you said, 40% said they used AI tools paid for by their employer to do this. And then 9% said they built their own tools for this purpose.

Like they intend to do it to build your own tool to do this means it's going to you plan on doing it a lot. That's a big commitment.

Well, or not, David, because you have a story about how it's getting easier and easier to build your own software. you spotted this zero YouTube series. I think it's called Is Everyone a Developer Now? Tell me about this.

I lost the story. One second.

Let me drill down on zero. Here we go.

And we've been talking about this on the show before where uh it's getting easier and easier with AI now, not just to do complex spreadsheets and workpapers that we couldn't do before, but you can actually make your own apps. And I'm really glad you brought this to the show, David, because I just did a webinar this week with McKenzie Thodash Basis, who has an accounting firm that specializes in crypto. And she walked me through in the webinar on earmark webinars, how she's using AI to make custom apps. She uses Perplexity computer and she can build an app that she publishes on a website through Perplexity, not using any other tools. And it can do really cool things like access the transaction history of a blockchain wallet and then export that history into a CSV file that she can import into QuickBooks. And this is something that she used to have to use specialized tools and pay a lot of money for it. Now she can just do it herself. And this what you're going to talk about kind of sounds like that.

And the conversation I think has shifted. It used to be I mean I've seen this happen growing this this momentum over the last 12 to 18 months. I vibe coded a replacement for QuickBooks. I built my own accounting thing, but they were always in like Reddit chat rooms, kind of these one-offs. They were never really a story story. But now I think it's going mainstream. And so the the in that's just habit of me saying because I've worked for intuitive developers, so it just comes out. The zero developer team, they have a new YouTube series. It's called is everyone a developer now? And essentially the de developer evangelist at zero, the three of them get on and they vibe code a solution or an app. And I mean to some extent it's an old idea. This is what your job is a developer evangelist. You would show the APIs of your app and show some cool thing you built on the APIs to get other developers excited to code more. So instead of them chasing a small pool of developers to build apps by having showing people how to vibe code, they basically now have opened up millions of accountants that could actually create apps and this hopefully would lead to more apps in the app store etc. So in their latest episode and I and I recommend everybody watch this and check it out because it'll give you a feel. Don't get hung up in the code details, but you'll get a feel of the thought process behind this where you use one AI to kind of create your plan and what your strategy is going to be. And then after you design it with the one AI, it gives you the prompt to paste into cloud code to go code the whole thing. And essentially that's what they do. So in an hour and 15 minutes, they coded up their own monthly close tool. And the each episode they do different tools like they did use lovable once a a fully hosted platform. Now they use cloud code in this one. Um, now it's only like halfway to be a production ready app. Like you couldn't vibe code this and then instantly share it with other people. But that's just, you know, a couple more weeks of iteration on the idea. But for to build an in-house tool, hour and 15 minutes.

So you said close month. I'm I'm thinking a month end close month type tool.

So it connects to zero, figures out if things have been reconciled. Here's your open invoices. Here's your open bills. Um, here's your balance sheet. Blah blah blah. So, it's

you can walk through that.

It has a UI and it's connected to zero. And you see it doing the work. And it's for an hour and 15 minutes. It's pretty impressive.

Yeah. Well, I believe it. Um, I actually used Cloud Co-work this week for another new task. I've been having so much fun with

Stay ahead of AI in accounting

Get the latest news on agentic AI for accounting, audit, and tax delivered to your inbox. Curated by AI, reviewed by professionals.

Subscribe to Newsletter