Myth: “The email came from our vendor’s own address, so the bank change must be real.”
It is a common belief, and payment-fraud attempts commonly lean on it. A matching sender address does not prove who is behind the message. Even when it comes from the real mailbox, the person typing may not be your vendor. And sometimes the address only looks real. Both cases can end with money sent to the wrong account.
Why a matching address proves little
There are two separate ways a familiar sender line misleads.

The address is genuine, but the sender is not. If an attacker has taken over a mailbox at your vendor, or at your own company, the mail is real mail from that mailbox. It can sit inside an existing conversation, with the usual signature, the usual tone and the right invoice references. Hidden inbox rules can also keep the real owner from seeing replies, so nobody on the other side notices the conversation has been hijacked.
The address only looks real. A forged From address is possible when SPF, DKIM and DMARC are missing or not enforced on the sending domain. Lookalike domains, with a swapped letter or an extra word, can also pass a quick glance, especially on a phone screen.
Email authentication results can help. A failed check is a useful warning sign, and your mail platform may flag it. But a pass does not settle the question. A lookalike domain can pass its own checks, and a compromised real mailbox passes them naturally. Treat the results as one input, not a verdict.
What actually holds up
A request to change bank details is a change to your payment process. It needs its own verification, whatever the sender line says.
- Call back on a number you already hold. Use your own records or an earlier contract, not the phone number in the email or its signature.
- Require a second approver. One person should not be able to change payee details and release the payment alone.
- Hold the first payment to a changed account. A short delay to verify costs little compared with a recovery effort.
- Look at the thread, not only the sender. Unexpected changes in tone, timing, reply-to address or payment method deserve a second look.
- Make urgency a trigger to slow down. Pressure to act quickly is a common feature of these requests, so staff should be encouraged to pause, not to hurry.
If a payment has already gone out
Contact your bank straight away. Recall requests tend to work better when made quickly, and the bank can tell you what is possible.

In parallel, start containment. Reset the password on any account you think was involved, revoke active sessions, and review mailbox rules and OAuth app consents for anything you did not create. Preserve the relevant messages and headers as you go, so the evidence is there for later, but do not hold containment back to do it.
Tell the vendor through a channel you have verified. Their mailbox may be the one that was compromised, and they will want to know.
Where to start this week
Write the callback step into your payment procedure on a single page, name an owner, and brief the people who handle invoices. If your own domain does not yet have SPF, DKIM and an enforced DMARC policy, add that to the list too, so your customers are not the ones guessing which of your emails is real.
Message us to talk through your payment verification plan.
Related
Discover more from Core-Infosec
Subscribe to get the latest posts sent to your email.