Blog/Business fraud

Supplier and vendor impersonation fraud

How fraudsters insert themselves into a real payment relationship, the four places they get in, and the small set of controls that actually stop the payment.

CyberWatch AI21 November 2026 · 12 min read
A printed invoice on a desk beside a calculator and a desk telephone

The fraud that costs businesses the most money does not look like a fraud. It looks like an invoice for work that was genuinely done, from a supplier you have used for six years, in a thread you have been replying to all month, with an attachment formatted exactly like the last fourteen. Everything in it is true except fourteen digits.

Vendor impersonation fraud works by inserting a stranger into a relationship that already has momentum. The attacker does not have to persuade you to make a payment, because you were going to make it anyway. They only have to change where it goes, and then stay quiet.

Why this variant is so expensive

Most fraud fails on a question. Did we order this? Who authorized it? Is this supplier on our list? Vendor impersonation survives all three, because the answer to each is yes.

Three features make it the costliest pattern in the FBI's business email compromise data:

  • The amounts are already large and already approved. Nobody has to be convinced that a six-figure payment to a construction contractor is plausible, because it is scheduled.
  • Every internal control passes. The purchase order matches. The goods were received. The approval limits are respected. The only field that is wrong is one that no system validates.
  • Detection is delayed by weeks. From your side, the invoice is settled and the ledger is clean. The alarm comes when the real supplier calls about a missing payment, by which time the money is long gone and the conversation is about who bears the loss.

MITRE ATT&CK now tracks this outcome under financial theft, grouping the techniques that end in a diverted payment rather than in data loss. It is worth noting the categorization, because it reflects a shift: for many criminal groups, the objective is a bank transfer and the intrusion is only the means.

The four ways in

There are only four positions from which someone can change a payment instruction convincingly. Knowing which one you are defending against determines what the defense looks like.

1. The supplier's mailbox is compromised

This is the most damaging version and the hardest to see. An attacker gets into the supplier's email, usually through credential phishing, then does nothing visible for days or weeks. They search the mailbox for terms like invoice, remittance, payment and bank, which ATT&CK records as email collection, and they read the customer correspondence until they understand the billing cycle.

Often they create a rule that quietly moves messages containing those words to an obscure folder, a technique ATT&CK catalogs as email hiding rules, so the real account owner never sees the thread the attacker is now handling. Then, when a genuine invoice is due, they reply from the real address, inside the real thread, with new bank details.

From your inbox, there is nothing to find. SPF, DKIM and DMARC all pass, because the message genuinely came from the supplier's domain. The signature block is theirs. The writing style is theirs, because it was copied from the thread. Our vendor email compromise article covers this pattern in detail, and our email security guide explains why authentication cannot help here.

2. Your own mailbox is compromised

The mirror image. An attacker inside your accounts payable mailbox watches the supplier conversations from your side, then either intercepts an incoming invoice and forwards it on with edited details, or writes to a colleague as though the supplier had confirmed a change. Because the request originates inside your own domain, external sender warnings and impersonation rules see nothing unusual.

3. A lookalike domain

The cheap version, and still effective. The attacker registers a domain resembling the supplier's, with a transposed letter, an added hyphen, a doubled character or a different top-level domain, and sets up full email authentication for it. ATT&CK lists domain acquisition as a preparatory technique for exactly this.

They then either start a fresh conversation or copy an existing thread into a new message, which reproduces the entire history underneath a slightly wrong address. On a phone, where only the display name is visible, there is nothing to notice at all. See lookalike domains for how to spot and monitor them.

4. The document or the side channel

Sometimes the email is genuine and the attachment is not. A compromised mailbox lets an attacker replace the PDF on a real invoice with an edited copy, keeping everything else identical. Other approaches move the conversation off email entirely: a message on a messaging app from a number claiming to be the supplier's accounts contact, or a phone call from someone who knows the invoice number and the amount because they have been reading the correspondence. Our articles on supplier bank change requests on WhatsApp and invoice documents sent by messaging app cover that route.

How the fraud unfolds

The sequence is consistent enough to be worth laying out, because recognizing where you are in it is sometimes the only warning available.

Reconnaissance. The attacker establishes who buys from whom. Your own website, case studies, tender awards, filings and staff profiles do most of this work. Job adverts for an accounts payable role will often name the systems you use and describe your approval process.

Access. One mailbox, on either side. Smaller suppliers are frequently the easier target, which is why a well-defended company is often attacked through a two-person firm that invoices it monthly.

Patience. This is the part defenders underestimate. The attacker waits for the right invoice, often weeks, because the fraud works best when the amount is unremarkable and the timing is expected. They learn who approves what, and they learn the one thing that matters most: whether you ever telephone to confirm changes.

The insertion. A short, calm, professional message. "Please note our banking details have changed following a group restructure. Updated details are on the attached invoice. Kindly confirm once your records are updated." No urgency, no threat, no spelling errors. Frequently accompanied by a plausible document: a letter on supplier letterhead, sometimes a fabricated bank confirmation.

The soft close. A follow-up a few days later, politely asking whether the update was processed. This is often the only pressure applied, and it is deliberately mild, because pressure invites scrutiny.

The payment and the silence. You pay. The attacker moves the funds through accounts and out, usually within hours. Nobody chases you, because from your records the invoice is settled.

The reveal. Weeks later, the real supplier asks about an overdue payment. Our article on invoices with changed bank details covers the first moves from that point.

Signals worth training people to notice

Detection by inspection is unreliable, and this should not be your primary defense. But some signals are genuinely worth knowing, particularly for the people who handle supplier records.

  • Any request to change bank details, at all. Treat this as the single trigger condition. Genuine changes do happen, which is exactly why fraudsters use them, so the response is verification rather than suspicion.
  • A reply-to address that differs from the sender, or a thread where the address changes subtly partway through.
  • A new individual introduced into an established relationship, especially someone described as handling accounts while a familiar contact is away.
  • An invoice identical to previous ones except for the remittance block. Compare against the last one you paid rather than against your idea of what it should look like.
  • An account in a different country, or a name that does not match the supplier. A payment to a third-party or factoring account may be legitimate, and it needs confirming through the supplier rather than through the person asking.
  • A request to move the conversation to a different email address, phone number or messaging app.
  • Pressure applied after your policy is explained. A genuine supplier accepts a verification call. Someone who objects strenuously to a twenty-four hour hold is telling you something.
  • A supplier who has just disclosed a breach. Expect fraud attempts referencing the relationship within days, as covered in our third-party risk guide.

The controls that actually stop it

The useful property of these controls is that none of them require anybody to identify the email as fraudulent.

Callback on a number you already held

Before any new bank details are used, call the supplier on a number from your own records, obtained independently of the request. Not a number in the email, the invoice, the attached letter or the signature block. Speak to a named person, state what you are verifying, and record who you spoke to and when.

This is the control, and everything else on this list supports it. The IC3 makes the same recommendation: verify changes to account information through a secondary channel. An attacker can control the email, the documents and the phone number printed on them. They cannot answer a line you already had.

Fix the onboarding gap first

Here is the reason callbacks fail in practice. When a supplier is first set up, the only contact details on file are usually the ones that arrived in the supplier's own email. A verification call later therefore goes straight to whoever sent the request.

Capture an independently sourced phone number at onboarding, from the supplier's published website or the signed contract, record who verified it and when, and make the field mandatory. Then audit your existing supplier master file for records with no verified number, because those are the accounts a fraudster will find easiest to change. This single piece of data hygiene is the difference between a verification policy and a verification capability.

Lock down the supplier master file

  • Changes to bank details made only by a small named group, never by the person who received the request.
  • Every change logged with who requested it, who verified it, how, and when.
  • An automatic notification of any change to someone other than the person who made it, and to the relationship owner.
  • Separation of duties so nobody can both change a payee and release a payment to it. See dual approval on payments.

A short hold on the first payment to a changed account

Twenty-four hours before the first payment to new or changed details costs almost nothing and removes the only pressure the fraud applies. A surprising number of these frauds collapse during that window, because the real supplier gets in touch about something unrelated.

Use whatever the bank offers

Ask your bank about account name verification, beneficiary confirmation or positive-pay style services and enable them. Where a payee name can be checked against the account name before the transfer, an obvious mismatch becomes visible before the money moves. Treat it as a useful layer rather than a substitute for the callback, since fraudsters open accounts in names chosen to pass.

Tell your suppliers how you work

Write to your suppliers once, stating that you will never change bank details on the basis of an email, that any change will be verified by a call to a number you hold, and that you expect them to verify changes from you the same way. This does two things: it sets the expectation so a genuine supplier is not surprised by the call, and it means an attacker inside their mailbox will read a description of a process they cannot beat.

Protect the mailboxes on both sides of the relationship

Phishing-resistant multi-factor authentication on finance, accounts payable and procurement mailboxes, alerting on new forwarding and message-hiding rules, and blocking of legacy authentication protocols. These prevent the position from which the fraud is run. For critical suppliers, it is reasonable to ask what they have in place, which belongs in the assessment covered by our third-party risk guide.

Write it down in one paragraph

The rule has to be short enough to be remembered under pressure, endorsed by leadership in their own words, and explicitly without exception for urgency or seniority. Our payment verification procedure article has wording you can adapt.

Got a request to change a supplier's bank details? Paste the message into CyberWatch AI Scan for a free check, then call the supplier on a number you already hold before changing anything.

If a payment has already gone

The first few hours decide whether anything is recovered. Work in parallel.

  1. Call your bank's fraud line immediately, not general customer service. Ask them to attempt a recall and to request a hold harmless letter or letter of indemnity from the receiving bank. Get a reference number.
  2. Report it to your national fraud or cybercrime body. In the United States that is the IC3, which asks for full banking details because they feed the recovery process. Elsewhere, report to the national police cybercrime unit or fraud reporting service.
  3. Freeze all further payments to that supplier and check whether other payees were changed in the same period. Attackers frequently alter more than one record.
  4. Call the real supplier on a known number. Their mailbox may be compromised, their other customers may be at risk, and you still owe the genuine debt.
  5. Secure your own mailboxes if there is any chance the compromise was on your side. Reset passwords, revoke sessions and tokens, remove forwarding and hiding rules and malicious app consents, and review the audit log for what was read.
  6. Preserve evidence before anyone tidies up. Full message headers, the original emails and attachments, the change record in your finance system, payment confirmations and a timeline.
  7. Notify your insurer within the policy window, even if cover is uncertain. Many cyber policies exclude fraudulent funds transfer unless crime or social engineering cover was added.
  8. Assess data exposure. If a mailbox was accessible, assume its contents were read, and consider whether that triggers any notification obligation.

Our articles on responding to business payment fraud and stopping a transfer quickly cover the sequence in more detail, and cross-border payment fraud covers the complications when the account is overseas.

Testing whether your controls hold

Policies fail quietly. Two cheap tests tell you most of what you need:

  • Pick three suppliers at random and ask the accounts payable team to show you the number they would call. Then check where that number came from. If it came from the supplier's email signature, you have a documented policy and no control.
  • Count how many new or changed payees in the last quarter have a documented verification. That percentage is your actual protection level, and it is usually lower than anyone expects.

Beyond that, walk a real bank-change request through with the people who would receive one, and run realistic practice emails against the finance team specifically rather than generic password-reset tests. Our payment fraud training article covers doing this without eroding trust.

The short version

Vendor impersonation fraud is not a technology failure. It is a correctly processed payment to the wrong account, arranged by someone who read your correspondence until they understood your billing cycle. No email check catches it, because the message is often genuine and the invoice usually is.

What stops it is dull and reliable: an independently sourced phone number captured when the supplier is onboarded, a mandatory callback to that number before any bank detail change takes effect, two people involved in creating and paying a payee, a short hold on the first payment to a new account, and leadership saying out loud that nobody will be criticized for delaying a payment to check.

For the wider picture, see our guides to business email compromise, vendor and supplier risk and business cybersecurity.

Frequently asked questions

What is vendor impersonation fraud?

A fraud in which someone poses as a supplier you already use and redirects a genuine payment to an account they control. The goods, the invoice and the contract are all real. Usually only the bank details have been changed.

How is it different from a fake invoice scam?

A fake invoice asks you to pay for something you never ordered, and it fails when someone checks the purchase order. Vendor impersonation uses a real debt you genuinely owe, so every internal check passes and only the destination account is wrong.

How do fraudsters know who our suppliers are?

From your own published material and theirs: case studies, press releases, tender awards, filings, job adverts and staff profiles. If a supplier mailbox is compromised, the attacker simply reads the customer correspondence.

Why do these frauds take so long to detect?

Because nothing looks wrong internally. The invoice is paid, the ledger balances and the purchase order matches. Discovery usually comes weeks later when the real supplier chases a payment you believe you already made.

What single control stops it?

A callback to a phone number you held before the request arrived, made before any new or changed bank details are used, with no exception for urgency or seniority. Everything else supports that one step.

Can we rely on the bank to catch a wrong account?

Only partly. Where account name checking services exist they will flag an obvious mismatch, which is genuinely useful, but fraudsters often open accounts in names close enough to pass. Treat it as one more layer, not the answer.

What if a supplier insists the change is urgent?

Urgency is the pressure the fraud depends on. A genuine supplier will accept a short delay for verification, and if they escalate aggressively against a documented policy, that is itself a warning sign worth escalating internally.

Is the money recoverable?

Sometimes, if you act within hours. Call your bank's fraud line immediately to request a recall, report to your national fraud or cybercrime body, and expect recovery to become unlikely once funds are withdrawn or moved abroad.

Should we tell the real supplier?

Yes, by phone, as soon as you know. Their mailbox may be compromised, their other customers may be targeted, and you will need their cooperation to establish what happened and to settle the genuine debt.

Sources

  1. Business Email Compromise (BEC), FBI Internet Crime Complaint Center
  2. Internet Crime Report 2024, FBI Internet Crime Complaint Center
  3. Industry Alerts, FBI Internet Crime Complaint Center
  4. Financial Theft, Technique T1657 - Enterprise, MITRE ATT&CK
  5. Acquire Infrastructure: Domains, Sub-technique T1583.001 - Enterprise, MITRE ATT&CK
  6. Hide Artifacts: Email Hiding Rules, Sub-technique T1564.008 - Enterprise, MITRE ATT&CK
  7. Email Collection, Technique T1114 - Enterprise, MITRE ATT&CK
  8. Phishing attacks: defending your organisation, UK National Cyber Security Centre
  9. Supply chain security guidance, UK National Cyber Security Centre
  10. Avoiding Social Engineering and Phishing Attacks, CISA
Share this article
LinkedIn WhatsApp X
For organizations

Your people get these messages at work too.

CyberWatch AI sends your staff realistic practice attacks, trains the gaps it finds, and shows management exactly where the organization stands.