• 11 Kasım 2026
  • Cyber Security

Corporate Email Security: Account Takeover and Invoice Fraud

Corporate Email Security: Account Takeover and Invoice Fraud

Introduction

  • Organisations spend most of their security budget on firewalls, endpoint protection and penetration testing. Yet the door attackers actually use is often far simpler: an email account.
  • The reason is straightforward. An email account is both an identity and a communication channel. Once an attacker is inside it, they do not merely read correspondence; they acquire the authority to speak in the organisation's name. No software vulnerability is needed, because the system recognises them as a legitimate user.

Passwords are not cracked, they are known

  • Account takeover is usually imagined as a brute-force attack: thousands of guesses until one works. In a significant share of real incidents, however, there is not a single failed attempt. The attacker signs in correctly on the first try.
  • That means the password was never cracked; it was already known. Three sources are common: a password leaked from another service and reused here, a password typed into a phishing page, or a saved browser credential harvested by infostealer malware on the user's device.
  • The distinction matters because it changes the response. If a password was cracked, a stronger password policy is the answer. If it was leaked, the replacement password will leak the same way; the real work is cleaning the user's device and their other accounts.

How invoice fraud (BEC) works

  • The most profitable use of a compromised corporate account is not sending spam but redirecting money. This attack class is known as BEC, or Business Email Compromise.
  • The pattern is usually this: the attacker reads past correspondence, finds an ongoing payment or order thread, then inserts themselves into it. They send a message that closely resembles a genuine invoice but carries altered bank details.
  • One technical detail deserves attention: the message genuinely originates from the organisation's own server and real address. SPF, DKIM and DMARC checks all pass, because the sender is not forged. The only forgery is intent.
  • To route replies back to themselves, attackers typically set the Reply-To header to a lookalike domain. When the recipient clicks "Reply", the message goes to the attacker rather than the organisation, and the exchange continues without either party noticing.

Why it goes unnoticed

  • In most organisations inbound mail is filtered heavily while outbound mail is not monitored at all. Yet the first sign of a compromised account appears in outbound traffic.
  • The second gap is the absence of sending limits. Without an hourly ceiling per account, a single mailbox can send thousands of messages for days, and nothing about it looks abnormal to the system.
  • The third is the noise created by undeliverable mail. Hundreds of bounce messages are frequently mistaken for a technical fault; in reality they are the earliest and clearest signal of an attack in progress.

Layers of technical defence

  • SPF, DKIM and DMARC. Together these define who may send mail on behalf of your domain, preventing others from impersonating you. They carry an important limitation: if the attacker is inside a real account, these checks will not stop them, because the mail genuinely comes from an authorised source.
  • Configure DMARC completely. Publishing a p=reject policy while failing to put the DKIM record into DNS is a common mistake. Your mail then rests on SPF alone, and any forwarding that breaks SPF causes outright rejection.
  • Multi-factor authentication. This is the single most effective control for rendering a leaked password useless on its own.
  • Outbound limits and alerting. An hourly ceiling per mailbox cuts abuse off within minutes. Set the threshold by measuring your organisation's genuine peak rate; otherwise legitimate bulk sending is blocked too.
  • Auditing password reuse. The same password across multiple accounts turns a single leak into a mass compromise.

What to do during an incident

  • Cut off access first: change the account password and terminate active sessions. A password change alone may not suffice; any forwarding rule, auto-reply or application password left behind by the attacker must also be removed.
  • Stop fraudulent messages still queued before they are delivered. Archive them before deletion; those records are needed both for forensic review and for identifying affected correspondents.
  • Block the attacker's source address permanently and audit other accounts using the same password.
  • Check the server's IP reputation. Outbound spam can place your domain on blocklists and disrupt mail flow for the entire organisation.
  • Finally, inform the account owner. If the password leaked, the problem is not on the server but on the user's side; a new password issued without a device scan will be compromised again shortly.

Conclusion

  • Email security is not a state achieved by installing a spam filter. Correctly configured authentication records, monitored outbound traffic and preparedness for an account takeover scenario are all parts of the same whole.
  • These attacks target a relationship of trust rather than a software flaw. Defence therefore cannot stop at technical controls; process rules matter too, such as verifying any change of payment details through a second channel outside email.
  • At Helix Teknoloji we provide support in corporate mail infrastructure setup, SPF/DKIM/DMARC configuration and email security auditing. If you would like your current setup reviewed, please get in touch.