Security & Email

Business Email Compromise, Explained

August 8, 2026 9 min read Haven Team

The costliest email attack does not carry a virus. It carries a plausible request from someone you trust, at a moment when you have a reason to act fast. Business email compromise consistently ranks among the highest-loss cybercrime categories in the FBI's annual reports, with reported losses in the billions of dollars, and it works because email was never built to prove who is writing to you.


Ransomware gets the headlines because it is loud. It locks screens, posts on leak sites, and shuts down hospitals. Business email compromise, usually shortened to BEC, is the opposite. It is quiet, it leaves almost no technical footprint, and it often succeeds on the first try. There is no payload to detect, no malicious attachment for a scanner to flag, no unusual process on the endpoint. There is just a message, and the message asks a person to do something they are authorized to do, worded so it feels routine.

That is what makes it hard. Most security spending assumes the threat is code. BEC is a con, and the target is a human decision, not a machine.

The Five Common Variants

The FBI groups BEC into a handful of recurring shapes. They share a structure but differ in who is impersonated and what the ask is.

Why the Supplier Variant Is So Effective

Invoice fraud deserves its own look because it defeats the instinct people rely on. Employees are trained to distrust the unexpected. A supplier-fraud email is the expected. Your company genuinely owes this vendor, the invoice number matches, the amount is right, and the only anomaly is a line that says the vendor changed banks. Vendors do change banks. The request is not implausible, it is mundane, and mundane requests do not trigger the alarm that a strange one would.

The attacker usually earns this precision through reconnaissance. Sometimes they sit inside a compromised mailbox for weeks, reading the thread between buyer and supplier, learning the tone, the names, the payment cadence, and they strike at exactly the right point in a real conversation. This is why account compromise and invoice fraud so often appear together.

The core weakness

Standard email carries no cryptographic proof of who wrote a message. The display name is a free-text field. The From address can be forged or merely look similar. When a request arrives, the recipient is authenticating the sender with judgment and familiarity, not with any technical guarantee.

How the Impersonation Is Done

There are three technical tiers, and it helps to know which one you are looking at.

Display-name spoofing is the crudest. The attacker sets the sender name to "Jordan Chen, CFO" while the actual address is some throwaway account. On a phone, many mail clients show only the name, so the address never appears. This costs nothing and still works.

Lookalike domains are the middle tier. The attacker registers a domain that reads like yours at a glance: a swapped letter, an added word, a different top-level domain. Replies go to the attacker, and the human eye skims right over the difference. Some of these use homograph tricks with characters that render almost identically to Latin letters.

Genuine spoofing and account takeover are the top tier. True From-address spoofing is what email authentication was designed to stop, and where SPF, DKIM, and DMARC are correctly deployed, forging your exact domain is much harder than it used to be. That is real progress. It is also why attackers have shifted toward lookalike domains and toward taking over real accounts, both of which sail past domain authentication because, technically, the mail is legitimate.

What Email Authentication Does and Does Not Cover

SPF, DKIM, and DMARC are worth deploying and every organization should. But be precise about the boundary. They verify that a message claiming to be from your domain was actually authorized to send as your domain. They do nothing about a message from your-company-invoices.com, which is a different domain the attacker legitimately controls, and nothing about a message from a real employee's hijacked account, which passes every check because it is authentic.

Attack Blocked by SPF/DKIM/DMARC?
Forging your exact domain Yes, when correctly enforced
Display-name spoofing No, the name is free text
Lookalike domain No, it is a real domain the attacker owns
Compromised real account No, the mail is genuine

The lesson is not that authentication is pointless. It closes the easiest attack and forces the adversary onto harder, more detectable ground. The lesson is that it is a floor, not a solution, and a control that stops only one of four variants cannot be the whole defense.

The Defenses That Actually Work

Because BEC targets a decision, the durable defenses are process controls, not filters. The single most effective one is out-of-band verification: any change to payment details, and any unusual transfer request, is confirmed through a separate channel using a number you already had, never the number or reply address in the email itself. A thirty-second phone call to a known contact breaks nearly the entire class.

Beyond that, a short list carries most of the weight:

If money has already moved, speed matters more than anything. Reporting a fraudulent wire to the bank and to law enforcement within the first day or two gives the best chance of a recall or freeze. After that window the funds are usually layered through mule accounts and gone.

Where Haven Fits

BEC exploits a gap Haven takes seriously: ordinary email carries no proof of who is really speaking. Encryption alone does not close that gap, which is why the honest framing is about verification, not just secrecy. When messages are cryptographically signed by keys tied to an identity you have confirmed, "is this really from them" stops being a judgment call and becomes something you can check. Haven signs and encrypts messages between users so provenance is a property of the message, not a guess based on a display name. It will not repair the wider email world overnight, and no tool substitutes for a callback before you move money. But the direction is the right one: make who-sent-this a fact the recipient can verify rather than a con the sender can exploit.

Try Haven free for 15 days

Encrypted email and chat in one app. No credit card required.

Get Started →