What account takeover is
Account takeover is a situation in which an outsider signs in to an employee’s account: the mailbox, Microsoft 365 or Google Workspace, the accounting system, the business bank account or an admin account on a server. As far as the systems are concerned, the person signing in is the employee. The password is correct, and sometimes the second verification step passed as well.
That is what sets it apart from other attacks. There is no suspicious file and no software encrypting anything. The attacker reads email, looks for invoices and conversations about payments, and uses the account to send messages in the organization’s name. Payment fraud in the name of a manager or a supplier (BEC, Business Email Compromise) often starts exactly this way: from a real account, inside a real email thread.
What raises the risk
- A reused password (password reuse): the same password on the work account and on another site that leaked. Lists of passwords from earlier leaks are used for automated sign-in attempts against email and cloud services (credential stuffing).
- An account without two-factor authentication (MFA): the password alone is all that stands in the way.
- A fake sign-in page: a phishing message that leads to a page that looks like the organization’s sign-in page and collects the details.
- A password stealer (infostealer): malware on a computer, sometimes the employee’s home computer, that pulls saved passwords and sign-in data from the browser.
- A flood of verification requests (MFA fatigue): an employee who gets approval requests on the phone again and again, and taps “Approve” just to make it stop.
- Forgotten accounts: the account of an employee who left and was never closed (offboarding), a shared account whose password everyone knows, or a supplier account that nobody keeps track of.
Early warning signs
The signs usually show up in small details, so they are only discovered when someone looks at them:
- A sign-in from a country, address or device that is not familiar, or at an hour when the employee does not work.
- Two sign-ins to the same account from distant places, with a time gap that makes traveling between them impossible (impossible travel).
- New mail rules (inbox rules) that the employee did not create: automatic forwarding to an external address, or moving certain messages to a folder nobody opens.
- Customers or suppliers who mention a strange message they received from the employee.
- Notifications about a password reset or a change of details that the employee did not initiate.
- Verification requests on the phone that arrive when the employee did not try to sign in.
Most business systems, including Microsoft 365 and Google Workspace, keep sign-in logs and can alert on some of these cases. The alerts help only if they are turned on and reach someone who actually handles them.
The impact on the organization
The familiar damage is financial: a message in the organization’s name asking a customer to send payment to a new bank account, or a request from a manager’s account to approve an urgent transfer. The message is sent from the real address, inside a conversation that is already going on, so it is very hard to spot.
A mailbox usually holds years of information: customer documents, copies of ID cards, contracts and payment details. Access to it is an exposure of personal data, which raises the question of whether there is a duty to report to the Privacy Protection Authority. A compromised account is also a springboard. Phishing messages go out from it to the contacts, it can lead to other systems that sign in with the same identity (SSO), and sometimes to admin accounts as well. And even when there is no financial loss, a customer who received a fraudulent message from the organization’s address remembers it.
How to reduce the risk
- Two-factor authentication (MFA) everywhere: email, remote access, financial systems and every admin account, including supplier accounts. On sensitive accounts, prefer methods that hold up against fake sign-in pages (phishing-resistant MFA), such as a physical security key or a passkey.
- Conditional access policy: blocking sign-ins from countries the organization does not work in, or requiring a managed device (MDM) to access sensitive data.
- A password manager: a long, different password for every service, without employees having to remember them.
- Identity and access management (IAM): a personal account for every person, permissions by role, and immediate closing of accounts when an employee leaves.
- Privileged account management (PAM): admin accounts kept separate from everyday work accounts and used only when needed.
- Sign-in monitoring: alerts on unusual sign-ins and on new mail rules, and someone who reviews them.
- A payment verification procedure: every change to bank account details is confirmed by a phone call to a known number, not to the number that appears in the message.
When an account is compromised: what to do right away
The first goal is to get the attacker out of the account and make sure there is no way back in. Changing the password alone is not enough, because the attacker may still have an active session or a permission they added themselves:
- Reset the password from a clean computer, and sign out all active sessions of the account (revoke sessions).
- Go over the account’s authentication methods and remove any device or phone number the employee does not recognize.
- Check mail rules, automatic forwarding and apps that were granted access to the account (OAuth consent), and remove anything unfamiliar.
- Inform the accounting team and stop recently approved payments that were based on messages from the account.
- Consider notifying contacts who received messages from the account, so they do not act on them.
If money has already been transferred, contact the bank immediately. The earlier you reach out, the more options there are to try to stop the transfer.
What the incident response (IR) team checks
Even when the account seems clean, it is worth having an incident response team look at the full picture. The answers they reach decide how the case is handled from there:
- How access was gained: a leaked password, a fake page or malware on the employee’s computer. If it was malware, the work expands to the whole computer and to every account that was used on it.
- Since when there was access and what was done with it, based on the sign-in and activity logs (audit logs).
- Which other systems the attacker reached, and whether other accounts in the organization were compromised.
- What information was exposed, and whether personal data is involved that requires considering a report to the Privacy Protection Authority.
At the end of the process the team recommends the fix that closes the gap: MFA on an account that was not protected, a conditional access policy, or a revised payment procedure. Sign-in logs are kept for a limited time, so reaching out early also matters so that there is something left to investigate.
Summary
- In an account takeover the attacker acts in the name of a real employee, which makes it hard to detect.
- The risk grows with reused passwords, accounts without MFA and accounts forgotten after someone left.
- The main signs are unusual sign-ins, mail rules nobody created and strange inquiries from customers.
- When it happens, sign out the sessions, check permissions and rules, stop payments and bring in the IR team.