Fraud

They sent your client a fake invoice, in your name

The email looked like yours. The logo was yours, the salesperson's name was right, even the signature. Only the bank account had been changed. Your client paid, and the money never reached your company.

July 28, 2026

It is one of the most profitable frauds there is, and it requires hacking nothing. They did not get into your email, they did not steal your password, there is no malware. They simply took advantage of something email has carried since it was designed: nobody checks the sender unless you demand it.

How it works, specifically

Writing an email that claims to come from ventas@[your-domain].com is as easy as writing that return address on an envelope. The system allows it. The only thing that can prevent it is your domain having told the world's servers what to do with email that isn't legitimate.

If your domain does not publish that instruction, or publishes it in observation mode, the forged email lands in your client's inbox exactly like one of yours. It does not go to spam. It carries no warning. It arrives clean.

The rest of the fraud is patience. They usually wait until there is a deal in progress — often because they have already been reading email from one of the two parties — and send the invoice exactly when one is expected. They change one thing: the bank details. Sometimes they add "we've changed banks, sorry for the inconvenience".

Your client was not careless. They received an email their server validated as authentic, because your domain gave it no reason to doubt.

Why it hits harder in Mexico

Because invoicing runs on email. Sending the PDF and the XML to the client, with the payment details in the body, is so routine that nobody questions it. An email with an invoice attached raises no suspicion at all: it is exactly what you expect to receive.

And the damage does not stop at the money. Your client paid in good faith to someone they believed was you, and now there is an awkward conversation about who absorbs the loss. That conversation costs the relationship more often than it costs the amount.

What does block it

A DMARC policy with teeth. DMARC is the record where your domain declares what a server should do when it receives an email claiming to be from you that fails the checks. It has three levels:

PolicyWhat happens to the fake invoice
p=noneIt reaches your client's inbox. You get told by report, afterwards.
p=quarantineIt goes to spam. Your client can find it if they go looking.
p=rejectIt never arrives. The server rejects it before delivery.

The difference between the first row and the last is the difference between finding out about the fraud and the fraud not happening.

How many companies are genuinely protected

We looked up the policy published by ten large Mexican domains on 28 July 2026. Six went all the way to reject. One is on quarantine. Three are still in observation mode: telmex.com, grupobimbo.com and cemex.com. All ten publish a reporting address, which means all ten started the project — three did not finish it.

It is worth noting that both banks in the sample are on reject. Where impersonation costs money immediately and visibly, the work gets finished. The detail of the measurement is here.

Can your domain be impersonated today?

The checker tells you which policy your domain has and whether it genuinely blocks or merely observes. It takes fifteen seconds.

Check my domain

If it already happened to you

  1. Warn your clients through another channel, not by email. Phone or WhatsApp. If the attacker is still inside the conversation, a warning email confirms to them that you spotted it.
  2. Ask for the complete original email, with headers. That is where you see which server it actually came from, and it is what a report needs.
  3. Check whether they also got into a mailbox. If the attacker knew details of the deal, they didn't just impersonate the domain: they read email. Change passwords and switch on second-factor authentication.
  4. Lock the domain down with p=reject so it can't happen again. This is the step usually left pending after the scare, once the urgency has passed.
  5. Report it. It is fraud, and the email headers are technical evidence.

And if it hasn't happened yet

This is one of the few things in security that you fix once and it stays fixed. It doesn't require anyone to remember anything, it doesn't depend on your team being careful, and it doesn't undo itself.

The only part to handle carefully is the route: jumping straight to reject without knowing who legitimately sends in your name blocks your own invoices too. That is why it is done in stages, with the reports in hand.

Keep reading