Skip to content
email security

Invoice fraud: how it works, what stops it and what to do in the first hour

Invoice fraud explained for finance and IT leads: six common schemes, the control that stops each, and what to do in the first hour after a bad payment.

A frosted glass invoice page sealed with a cracked brass wax seal, with the headline Invoice fraud: how it works and how to stop it

Invoice fraud is any scheme that gets a company to pay a bill it doesn't owe, or to send a payment it does owe to a criminal's account. The controls that stop it are plain: a call-back to a known number before bank details change, a second approver, a hold on the first payment to any new account, and email authentication on your domains.

This guide is for finance and IT leads at companies with many suppliers, several approvers and more than one entity, and it runs during International Fraud Awareness Week, November 15 to 21. For the email records behind the last control, see our guide to DMARC policy not enabled.

Why does invoice fraud still work?

Because the payment looks routine. The Federal Trade Commission's guide to scams aimed at businesses puts it plainly: scammers "hope the person who pays your bills will assume the invoices are real and make the payment." With hundreds of invoices a week, one more bill from a known supplier rarely gets a second look.

The losses are large. The FBI's Internet Crime Complaint Center (IC3) counted 24,768 complaints of business email compromise in 2025, with $3,046,598,558 in reported losses, second only to investment fraud, in its 2025 annual report. It describes a scam aimed at those "working with suppliers and/or businesses regularly performing wire transfer payments," carried out by compromising email accounts and other channels, such as phone numbers and virtual meetings.

What are the common invoice fraud schemes?

Most cases follow one of six patterns, and no single control covers them all.

Scheme What it looks like The control that stops it Who owns it
Fake vendor invoice A bill for goods or services nobody ordered Pay only against a purchase order and a receipt Accounts payable
Bank-detail change "Our bank has changed," with new account details for the next payment A call-back to a number already on file, then a hold on the first payment Owner of the vendor file
Exact-domain spoofing Your address, or a supplier's, in the From line of mail neither of you sent DMARC at quarantine or reject on every domain IT
Look-alike domain An address one letter off, or the right name with a different ending An outside-mail banner, the full address shown, and the call-back IT and accounts payable
Compromised supplier mailbox A real email, in the real thread, from the supplier's own account The call-back and a second approver Finance
Compromised internal mailbox Your own executive or payables inbox asks for, or hides, a payment Multi-factor sign-in, no outside forwarding, logged rule changes IT

The supplier row matters more than it looks. The FBI's alert on business email compromise through cloud email describes criminals who read a compromised account for evidence of payments, then "impersonate email communications between compromised businesses and third parties, such as vendors or customers," asking for pending or future payments to go to their own accounts. It adds that one compromised mailbox can pivot to multiple victims within an industry.

What does email authentication stop, and what doesn't it?

Faking the From line is easy. NIST's guide to trustworthy email, SP 800-177 Rev. 1, calls spoofing the From address "trivial" and names domain-based checks (SPF, DKIM and DMARC) as "the first line of defense," able to "alert on or block email that was sent using a spoofed domain." CISA's Cybersecurity Performance Goals 2.0 recommend SPF, DKIM and DMARC on all corporate email, with DMARC set to reject.

Two limits matter for invoice fraud:

  • Your record protects others. Your DMARC policy tells your customers' and suppliers' inboxes what to do with mail faked in your name. Mail faked in a supplier's name depends on the supplier's record. So check your domain, then run your largest suppliers' domains through the same free check. Many domains are open: our October check found 79% of 519 South Florida law firms had nothing telling inboxes to block fakes in their name (see law firm email spoofing).
  • It can't stop a real mailbox or a near-miss address. NIST notes that these checks "will not prevent malicious email sent using compromised legitimate accounts." A look-alike domain belongs to the criminal, so your DMARC record has no say over it; NIST describes a From address that "looks very close to one of the legitimate addresses the user is familiar with."

Email authentication closes one door; the payment process closes the rest. Before you move to reject, the reports show every service that sends as your domain: see how to read DMARC reports.

Which payment controls should finance own?

Write these into the payment policy, for every entity and location.

Verify every bank-detail change by phone. Call a number you had before the request arrived, from the vendor file or the contract, never one from the request itself. The FBI's advice is to "verify all payment changes and transactions in person or via a known telephone number," and its 2024 update on business email compromise adds secondary channels or two-factor authentication. Place the call yourself: the FTC warns that "scammers often fake their phone numbers," so caller ID proves nothing.

Separate who changes the vendor file from who approves payment. CISA's goals call for separation of duties, "distributing responsibilities across multiple individuals or roles to reduce the risk of unauthorized actions, errors, or fraud." In practice, the person who enters a new bank account can't release a payment to it, and payments above a limit finance sets need a second approver. Urgent requests from executives get no exception.

Hold the first payment to any new or changed account. This one is our advice, not a rule from the sources: hold it until the call-back and the second approval are both on file. Tell suppliers the hold exists, so a real supplier expects it and a fake one can't rush it.

Pay only what was ordered and received. The FTC asks for clear procedures "for approving purchases and invoices" and for staff to check every invoice closely. Match each invoice to a purchase order and a receipt. It also says to "pay attention to how someone asks you to pay": a sudden switch from check to wire, or to a bank in a new country, is a reason to call.

What should IT change in email?

Most of this list comes from the FBI's cloud email alert:

  • Multi-factor sign-in on every mailbox, including shared payables inboxes.
  • No automatic forwarding to outside addresses. Criminals set mailbox rules that "delete key messages" and forward mail out.
  • Logs of mailbox sign-ins and settings changes, kept for at least 90 days.
  • A banner on mail from outside the company.
  • SPF, DKIM and DMARC on every domain, "to prevent spoofing and validate email."

The 2024 update adds two habits for staff: check the full sender address, especially on a phone, and watch for links "that may contain misspellings of the actual domain name." Our email security checklist puts these items in one list, with an owner for each.

What should you do in the first hour after a fraudulent payment?

Speed decides how much comes back. The IC3 report is blunt: "If you discover a fraudulent transfer, time is of the essence." In order:

  1. Finance calls the bank's fraud line and asks for a recall of the funds "along with any necessary indemnification documents." Policies differ by bank, the report notes, so learn yours before you need it.
  2. Finance files a complaint at ic3.gov, whatever the amount, with the full transaction details. The IC3 may be able to help banks and law enforcement freeze the funds. In 2025, its Recovery Asset Team worked 3,900 incidents of all kinds, with a 58% success rate.
  3. Accounts payable stops every other payment to that account and flags it in the vendor file.
  4. IT secures any mailbox that may be involved: reset the password, end open sessions, remove forwarding and delete rules, and keep the logs.
  5. Someone calls the real supplier on the number on file, to learn whose mailbox sent the request and to warn them.
  6. Everyone keeps the evidence: the email with its full headers, the invoice and the payment record. Delete nothing.

Later that day, tell your insurer and legal counsel, and report the scam to the FTC at ReportFraud.ftc.gov, as its guide asks.

How do you test your process this week?

Pull the last five bank-detail changes in the vendor file and ask three questions of each: who made the call-back, to which number, and who approved the first payment. Any answer that is "nobody" or "the number in the email" marks the control to fix first. Then ask IT which domains you own and what DMARC policy each one has.

How can Raileon help?

Our free email security check reads a domain's SPF, DKIM and DMARC records from public DNS and gives a plain answer: protected, partly protected or at risk. It sends no email and never touches your mail server. If the answer isn't protected, we set up SPF, DKIM and DMARC, then check your real email still gets delivered. And in the systems we build and run, payments always wait for a person, and that can't be switched off. See how we work.

FAQ

What is invoice fraud?

It is any scheme that gets a company to pay a bill it doesn't owe, or to send a payment it does owe to a criminal's account. Fake vendor invoices, bank-detail change requests and compromised supplier mailboxes are the common forms.

How does a fake invoice scam work?

A criminal sends a bill that looks like one your company expects, or asks for the next payment to go to a new account. It works when the invoice is paid without a matching order, or the bank change is accepted without a call to a known number.

Is invoice fraud the same as business email compromise?

They overlap. Business email compromise uses a compromised or faked email account to redirect payments, and many invoice frauds work that way, but a fake invoice from a made-up supplier needs no break-in at all.

Does DMARC stop invoice fraud?

Only one kind. DMARC at quarantine or reject tells inboxes to send mail faked with your exact domain to spam or refuse it, but it does nothing about look-alike domains or a real mailbox someone has broken into.

Is your payment process ready for a fake invoice?

Bring your vendor-change and approval policy to a free 20-minute call. We'll go through it with you and point out where a fake request could get through.

Book a free 20-minute call

About the author

Konstantin, Founder & CEO of Raileon

Konstantin

Founder & CEO, Raileon

Konstantin has spent more than 15 years in systems, processes, automation and development. He has built businesses from scratch, from early crypto to tech ventures that led to acquisitions, and founded Raileon to build the software and automation he wished he'd had.

  • 15+ years in IT, automation and development
  • Founder & CEO of Raileon
  • Based in North Miami, Florida

Let's put your operations on rails

Book a 20-minute call with our team in Miami, Florida, and we'll show you exactly what Raileon would automate for your business.

Book a 20-minute call

No commitment. No credit card. Just a conversation.