A customer calls to ask about the document you shared this morning. You didn't share anything. When you look in your Sent Items, there's nothing there, but your contacts have clearly received an email from your address with a link to a "secure file".
Your mailbox has been taken over, and the attacker is using your name and your reputation to phish the people who trust you most. That's what makes this attack work: a message from a real supplier, sometimes sent as a reply in an existing conversation, gets far more trust than a cold phishing email.
The good news is that the clean-up follows a clear order, and most of it can be done in an afternoon. This guide is for the organisation whose account was hacked: what to do today, what to check, and how to make sure it doesn't happen again.
Step 1: lock the attacker out
Changing the password on its own isn't enough. An attacker who is already signed in keeps a valid session, and Microsoft's guidance for cutting off access is to disable the account and revoke its sessions so every app has to ask for a new sign-in.
What to do:
- In the Microsoft 365 admin centre or the Microsoft Entra admin centre, block sign-in for the account while you work.
- Reset the password to a new, long password that isn't used anywhere else.
- Select Revoke sessions (sometimes shown as "Sign out of all sessions"). Apps that already hold an access token can keep it for up to an hour by default, so keep sign-in blocked until you've finished the next steps.
- If the same password was used on other sites, change it there too.
Step 2: check the MFA methods
Attackers often add their own sign-in method as soon as they get in: an authenticator app or a phone number they control. That lets them keep access even after the password is reset, because the next sign-in prompt goes to their device.
What to do:
- In the Entra admin centre, open the user's Authentication methods and remove anything the user doesn't recognise. If in doubt, remove all methods and have the user register again in person.
- Check the account's audit log for recent "security info" registrations. A new phone number or app added on the same day as the phishing is a strong sign.
Step 3: look for forwarding and inbox rules
This is the step people skip, and it's where attackers hide. Microsoft lists inbox rules the user didn't create as a typical sign of a compromised account, especially rules that forward mail to unknown addresses or move messages to folders like Notes, Junk Email or RSS Subscriptions. The point is to hide the replies: when a contact writes back "did you really send this?", the owner never sees it.
What to do:
- In Outlook (or as an admin in the Exchange admin centre), open the mailbox's rules and delete anything the user didn't create. Look for rules with short or odd names, or rules that act on words like "invoice", "payment" or "phishing".
- Check forwarding on the mailbox itself, not just in the rules.
- Look in the Deleted Items, Junk, RSS Subscriptions and Archive folders for replies from contacts that the user never saw.
- Check the user's list of connected apps and remove any they don't recognise. A malicious app with access to the mailbox survives a password reset.
Step 4: review the sign-in logs
The sign-in logs tell you when the attacker first got in, from where, and with which app. That matters because the phishing is often not the first thing they did: they may have been reading mail for days.
Look at the last few weeks for sign-ins from unusual countries, unknown devices, or apps the user doesn't use. Be aware that Microsoft Entra keeps sign-in logs for 7 days on the free tier and 30 days with Entra ID P1 or P2. P1 comes with Microsoft 365 Business Premium. On a basic licence, if you wait a week, the evidence is gone.
What to do:
- Export the sign-in logs for the affected account today, before they expire.
- Note the first suspicious sign-in and the IP addresses involved.
- Check whether the same IP addresses signed in to other accounts in your organisation. One compromised mailbox is often not the only one.
- Read through the Sent Items and the mailbox audit log to see what the attacker sent and to whom.
Step 5: warn your contacts
Your contacts are the real targets now. Tell them quickly and clearly, so nobody clicks the link or pays an invoice from the hijacked conversation.
What to do:
- Send a short message to everyone who received mail from the account in the affected period. Say what happened, when, what the fraudulent email looked like, and that they should delete it and not open links or attachments.
- Phone the contacts who handle payments with you. Tell them you won't change bank details by email, and that any such request should be checked by calling a number they already have.
- If someone already entered their password on the phishing page, they should treat their own account as compromised and follow these same steps.
- Forward a copy of the phishing email to suspicious@safeonweb.be, which helps the CCB warn others.
Don't send the warning from the compromised account until steps 1 to 3 are finished. Otherwise the attacker can see it, and may answer it.
Step 6: decide whether it's a data breach
An attacker with access to a mailbox can usually read everything in it: customer details, contracts, employee information. Under the GDPR, a personal data breach must be reported to the Data Protection Authority within 72 hours of becoming aware of it, unless it's unlikely to result in a risk to people. Our ransomware article explains the Belgian notification process. Write down what you found in steps 3 and 4: it's the basis for that decision, and for any insurance claim.
How to stop it happening again
Most mailbox takeovers start with a password typed into a convincing fake login page. These are the controls that make the biggest difference.
- Use MFA that can't be phished. Text codes and push approvals can be relayed by a fake login page. Passkeys can't. Our article on phishing-resistant MFA explains how to roll them out.
- Block automatic forwarding to external addresses unless there's a documented business need.
- Alert on new inbox rules, forwarding and new MFA registrations, so the next compromise is spotted in hours rather than when a customer calls.
- Upgrade the email protection. Phishing from a real, trusted partner passes the usual sender checks because it really does come from their mailbox. Detection that looks at behaviour and content, not just the sender, catches more of it.
- Make a written payment rule: bank details are never changed on the strength of an email alone.
- Train your team to report a suspicious email instead of deleting it, and to report a click straight away. A fast report is worth more than a perfect record.
A managed partner can watch for these warning signs across your mailboxes and clean up quickly when something gets through, but the basics above are worth doing whoever looks after your IT.
Your checklist
- Block sign-in, reset the password and revoke all sessions.
- Remove unknown MFA methods and connected apps.
- Delete forwarding and inbox rules the user didn't create.
- Export and review the sign-in logs for the last few weeks, and check other accounts.
- Warn your contacts, by phone for anyone who handles payments.
- Assess the GDPR question and keep notes of what you found.
The takeaway
A hacked mailbox is embarrassing, but the response is straightforward: lock the attacker out, remove what they left behind, check the logs and warn your contacts. Do it in that order and do it today, because the logs don't wait. Then spend an hour on the prevention list: phishing-resistant MFA and blocked external forwarding alone close most of the doors this attack uses.