Request a quote
Emergency SectorsCareers About us Blog Get in touch
Two colleagues at a table with a laptop between them in a warmly lit office

CEO fraud: the payment process that stops it

CEO fraud arrives as an email, but it only becomes a loss when someone pays. That is where the defence belongs: not in another session on spotting suspicious emails, but in a payment process that makes no exception for urgency, secrecy or a director. This is that process, and what IT does around it.

8 min read Written by the ITproposal team
Short answer

CEO fraud is not an email problem but a payment problem: the message may be believed, and the payment still must not go out.

Five rules cover the known variants: no payment outside the normal route, a call back to a known number, a second approver outside the request, a limit enforced in the banking platform itself, and secrecy as a reason to stop.

If the money has gone, call your bank at once. Only the bank can recover it, and an investigation that stands up in court is work for a forensic firm; we make sure the evidence is still there.

What CEO fraud is, and why it is not an email problem

In CEO fraud, someone poses as your managing director, a board member or a regular supplier and asks a member of staff to transfer money. The request is urgent, often confidential (“there’s an acquisition under way, don’t mention it to anyone yet”) and arrives when checking is awkward: late on a Friday, during the holidays, while the director is travelling. It needs no malware and no vulnerability. The attacker uses the authority your own employee already has.

Three forms are the most common:

  • The request from the boss: an email or chat message from an address that resembles the director’s, or from their real mailbox after it has been taken over.
  • The changed bank details: a “supplier” reports a new account, often from a mailbox that genuinely belongs to that supplier and has been compromised. The invoice is right; only the account number is not. In the UK this is usually called payment diversion fraud.
  • The voice on the phone: a call or voice message, increasingly with a cloned voice, that “confirms” an earlier email.

How to deal with the email and the click is covered in our article on phishing and the human layer. This article is about what comes after: once the message is in and believed, the payment should still be unable to leave.

The payment process that stops CEO fraud

A procedure only works if it is short enough to remember and strict enough to leave no room. These five rules cover the known variants:

  • No payment outside the normal route. A payment comes from the purchasing or invoice process, with an order or a booked invoice behind it. A request by email, chat or phone is a request, not an instruction, even when it comes from the director.
  • Call back on a known number. Any unusual request is checked using a number you already had: from the staff list, the contract or the supplier records. Never the number or link in the message, and never by replying to the email.
  • Dual approval, with the second approver outside the request. The second sign-off comes from someone the requester has not contacted. Two people who received the same message are one check, not two.
  • A limit above which two signatures are always required. Set it in the banking platform itself, not only on paper; many business banking platforms can enforce a second authorisation.
  • Secrecy is a reason to stop. A request that must not be discussed with the finance lead goes to the finance lead first.

Put the procedure on one page, have the board sign it and record who may change it. A rule that lives only in the head of the person who runs accounts payable disappears the first time they go on leave.

New bank details are a process of their own

The most expensive variant is not an urgent request but a change to supplier master data. Whoever gets a regular supplier’s account number changed receives every invoice payment after that, and nobody notices until the real supplier rings about arrears.

  • Never change bank details on the strength of an email or an invoice, even if the email comes from the right address. That address may be compromised.
  • Confirm with a contact you already knew, through a channel you already had. Do not ask for confirmation on the number given in the email.
  • Have two people record the change: one enters it, the other checks it. The finance system should log a change of bank details with a name and a time.
  • Build in a waiting period. The first payment to a new account goes only after the check, and preferably not on the same day.
  • Take your bank’s name check seriously. UK banks check whether the name matches the account (Confirmation of Payee), and banks in the euro area have done the same since October 2025. A “no match” warning is often clicked past in practice; make it a rule that the warning stops the payment until someone has called back.

The rule applies to the board too

Most organisations have a rule like this and quietly skip it when the request comes from someone senior. That is exactly what CEO fraud counts on: an employee who would rather take a risk than contradict the director.

Turn it round. Have the director state, in writing and in advance, that they will never ask anyone to skip the procedure, and that nobody who calls back or refuses will ever be criticised for it. Refusing then stops being disobedience and becomes following an instruction from that same director.

Two additions that cost very little:

  • An agreed check question for the few people who may approve payments, written nowhere in email or chat.
  • A deputy for each role. When the director or the financial controller is away, it is settled who approves instead. Otherwise “they can’t be reached” becomes the argument for skipping the check.

How to recognise CEO fraud

Recognition is the second line of defence, not the first: a good impersonation has no spelling mistakes. The shape of the request gives it away more often than its content.

  • Urgency with a reason not to check: “it has to go today”, “I’m in a meeting, don’t call”.
  • Secrecy: an acquisition, a fine, a confidential deal.
  • A new channel: the director suddenly getting in touch on WhatsApp or from a personal address.
  • An odd address: a domain one letter off, or a reply-to address that differs from the sender.
  • A new account number, often in a different country from the supplier.
  • A moment when checking is hard: Friday afternoon, a bank holiday, the first day back after leave.

None of these signs is proof on its own. The point of the procedure is that nobody has to judge: with any unusual request, you call back, even if it looks genuine.

What IT adds

Technology does not replace the payment process, but it makes the attack harder and the traces easier to see. These are the controls that matter most:

  • MFA and conditional access on every mailbox, and certainly on those of the board and finance. A compromised mailbox is the most convincing sender there is.
  • An alert on new forwarding and inbox rules. An attacker who is reading along first sets a rule that hides replies from the bank or the supplier.
  • A tag on external senders, so that an email “from the director” that comes from outside stands out.
  • SPF, DKIM and DMARC on your own domain, so that nobody else can quietly send email in your name.
  • Monitoring for lookalike domains, so that a registration one letter off is spotted before the first email goes out.
  • Dual approval in the banking platform and the finance system, as technical enforcement of the two-person rule.

Keep email and sign-in logs for long enough as well. If it does go wrong, the bank, the police and your insurer will want to know what happened, and that is only in logs that still exist.

If the money has already gone

Then every hour counts. Recovery is not guaranteed and depends on how quickly the receiving bank can freeze the funds.

  • Call your bank at once on its fraud or emergency line and ask for the payment to be recalled. Give the time, the recipient’s account details and the reference.
  • Report it. In England, Wales and Northern Ireland that is Action Fraud; in Scotland, Police Scotland. Your bank will usually ask for the reference.
  • Keep the evidence: the emails with their full headers, chat messages, call records. Delete nothing, and export emails rather than just forwarding them.
  • Secure the mailbox if it was taken over: revoke sessions, reset the password and MFA, check rules and connected apps.
  • Warn suppliers and customers if your mailbox was used; they are often sent the next email.

If personal data could be viewed through that mailbox, it may also be a personal data breach that you need to assess and possibly report.

Where our work stops

We are not a bank. Stopping, recalling or freezing a payment is something only your own bank can do, together with the receiving bank. We cannot do it and we do not promise it.

We are not forensic investigators. An investigation to establish who did it, or evidence that will stand up with a court or an insurer, is work for a specialist forensic firm. We make sure the logs exist, that a mailbox is secured without wiping traces, and that such a firm gets access.

What we do: the email and account security that makes the attack harder, enforcing dual approval in the systems we manage, and the first hour after an account takeover. The payment process itself belongs to your finance team and your board; we help write it down and check that the technology supports it.

Frequently asked

What we are asked about CEO fraud

The questions we hear, usually just after an attempt.

What is CEO fraud?

In CEO fraud, someone poses as the managing director, a board member or a regular supplier and asks an employee for a payment or a change of bank details. The request is usually urgent and confidential. No hacking is needed: the attacker uses the authority the employee already has.

How do you spot CEO fraud?

By the shape of the request: urgency, secrecy, a new channel such as WhatsApp, an odd email address or new bank details, and often a moment when checking is awkward. None of these is proof on its own, which is why any unusual request is checked by calling back on a number you already had.

How can a business prevent CEO fraud?

With a payment process that has no exceptions: no payment outside the normal route, a call back on a known number, a second approval by someone the requester has not contacted, and a separate check for every change of bank details. Have the board confirm in advance that the rule applies to them too. MFA and alerts on suspicious mailbox rules make the attack harder on top of that.

What should you do if a payment has already been made?

Call your bank on its fraud or emergency line at once and ask for a recall; whether the money comes back depends mostly on speed. Report it, keep the emails with their full headers and secure the mailbox if it was taken over. Warn suppliers and customers if your mailbox was used.

Can you recover the money or find out who did it?

No. Only your bank can recover a payment, and a forensic investigation that stands up with a court or an insurer is work for a specialist firm. We secure the mailbox, preserve the logs and give that firm access, and we set up the security and dual approval that make the next attempt harder.

Related services

Further reading

The technology around the payment process.

Is your payment route closed?

In half an hour we go through your email security and the approvals in your systems, and we say what belongs with your bank or your accountant.

Practical IT knowledge in your inbox

New guides on management, security and the workplace, written by the people doing the work. No sales talk, and you can unsubscribe in one click.

We use your address for the newsletter only. See our privacy policy.

Request a quote Call