How Invoice Fraud Starts With One Compromised Email Account
A controller at a 70-person distribution company pays a supplier's invoice on a Thursday afternoon. The invoice came from the supplier's real email address, in the same thread as last month's order, with a short note that the company had switched banks. Three weeks later the supplier calls to ask why they haven't been paid.
Nobody in that story was careless. The email came from a real account that someone else was controlling. This post covers how invoice fraud starts, what attackers do inside a mailbox, what it costs, and the controls that stop the payment, whether the compromised account is yours or a vendor's.
The scale of the problem
The FBI tracks this under business email compromise (BEC). In its 2024 Internet Crime Report, the FBI's Internet Crime Complaint Center (IC3) recorded 21,442 BEC complaints with about $2.77 billion in reported losses. That works out to roughly $129,000 per complaint on average. The 2025 report put BEC losses at about $3 billion, second only to investment fraud among all the crime types the FBI tracks.
Those are only the losses people reported. In a September 2024 public service announcement, Business Email Compromise: The $55 Billion Scam, the FBI counted more than $55 billion in exposed losses worldwide between October 2013 and December 2023.
For a company with 25 to 200 employees, one redirected payment of that size is a serious loss. Unlike ransomware, nothing breaks. Systems keep running, and the loss often surfaces weeks later.
How one account becomes a payment redirect
Invoice fraud is rarely a single fake email. It usually follows a pattern with four stages.
1. The attacker gets into a mailbox
The most common way in is a convincing sign-in page. Someone gets a "shared document" or "voicemail" email, clicks through, and types their password into a page that looks exactly like their normal Microsoft 365 or Google sign-in.
Multi-factor authentication stops many of these attempts, but not all of them. Some phishing kits sit between the user and the real sign-in page, pass the login through, and capture the session token that proves the user already passed their second factor. CISA described this kind of "pass-the-cookie" technique in its analysis report AR21-013A, where attackers reached cloud email accounts even though the user had signed in with MFA. Text-message and simple push-approval MFA are the most exposed. CISA's fact sheet on phishing-resistant MFA ranks SMS and voice codes as the weakest options and FIDO security keys or passkeys as the strongest.
2. The attacker hides and reads
Once inside, the attacker doesn't send anything right away. They read vendor threads, open invoices, payment schedules and the names of the people who approve payments.
To stay hidden, they create mailbox rules. A typical rule moves any email containing words like "invoice," "payment" or "bank" into an obscure folder, or forwards copies to an outside address. The FBI and CISA warned about this exact tactic in a 2020 private industry notification: criminals were setting auto-forwarding rules in web-based email, and those rules often didn't sync to the desktop email app, so neither the user nor IT saw them. Microsoft's own documentation on detecting malicious Outlook rules notes that these rules live in the cloud mailbox, so reinstalling Outlook or replacing the laptop doesn't remove them.
3. The attacker times the request
When a real invoice is due, the attacker replies inside the existing thread. The message comes from the real account, so it passes every sender check. It references the actual purchase order and asks for one small change: new bank details, a temporary account for "an audit," or an updated remittance form attached as a PDF.
Replies from the real vendor are quietly moved by the hidden rule, so nobody sees the conversation that would expose it.
4. The money moves, then moves again
Once the payment lands, it can be moved on to other accounts, sometimes at several banks, which makes it harder to trace and freeze. Speed matters on the victim's side too, which is covered below.
Two directions: your mailbox or theirs
The same attack runs in two directions, and growing companies are exposed to both.
Your account is compromised. The attacker uses your accounts payable clerk's mailbox to send your customers "updated" payment instructions, or uses your CFO's mailbox to tell your own finance team to pay a new vendor. Your customers lose money, and the trail leads back to your domain.
Your vendor's account is compromised. You do everything right internally, and the fraudulent invoice still arrives from a supplier you've paid for years. Their mailbox was the one taken over. From your side, the email is indistinguishable from the real thing.
This is why email filtering alone can't solve invoice fraud. A filter can block a lookalike domain. It cannot tell you that a genuine message from a genuine vendor was written by someone else.
What it costs beyond the payment
The redirected payment is the visible loss. Several others follow:
- Recovery is uncertain. The FBI's Recovery Asset Team works with banks to freeze fraudulent transfers through its Financial Fraud Kill Chain. Its 2025 report says it froze about $679 million of roughly $1.16 billion in attempted theft across about 3,900 incidents, a 58% success rate. The process depends on the victim's bank and the FBI hearing about the transfer quickly, as the FBI's domestic kill chain guidance explains. A payment discovered three weeks later is much harder to stop.
- The vendor still hasn't been paid. In the vendor-compromise scenario, the real invoice is still outstanding, and whether you owe it again depends on your contract terms.
- Insurance terms matter. Check whether your cyber policy covers social engineering or funds transfer fraud, at what limit, and under what conditions. Read the policy language before you need it.
- A mailbox breach can be a data breach. If the attacker had access to a mailbox full of customer or employee records, you may have notification obligations on top of the financial loss.
The controls that actually stop it
No single control prevents invoice fraud. The ones below work as layers: keep attackers out of mailboxes, spot them quickly when they get in, and make sure a changed bank account never gets paid on email alone.
Payment process controls
These cost almost nothing and stop the fraud even when every technical control fails.
- Call-back verification for any bank detail change. Call the vendor at a phone number already on file, never one from the email requesting the change. The FBI's BEC guidance recommends verifying account change requests through a secondary channel.
- Two-person approval for new payees and for changes to existing payee bank details.
- A short hold on first payments to newly changed accounts, so there's time for the real vendor to notice.
- Tell your customers, in writing, that you will never change your bank details by email. Put it on invoices.
Mailbox and sign-in controls
- Stronger MFA for finance and executive accounts. Move the people who approve or send payments to phishing-resistant methods such as security keys or passkeys, and turn on number matching for everyone else. Our post on MFA rollouts covers where these deployments usually leave gaps.
- Block automatic forwarding to outside addresses at the organization level, with documented exceptions.
- Turn off legacy sign-in protocols that can't use MFA.
- Conditional access rules that flag or block sign-ins from unusual countries or unmanaged devices.
Monitoring
This is where most growing companies fall short. The signs of a compromised mailbox usually show up in sign-in and mailbox audit logs: a sign-in from an unexpected location, a new inbox rule that hides finance emails, a burst of mailbox searches for "invoice." Someone has to be watching those signals and know what to do with them. Without that, the attacker gets weeks of quiet reading time.
What responding looks like
When a compromised mailbox is caught, the first hour follows a fairly standard sequence:
- Cut off access. Reset the password, revoke all active sessions and tokens, and re-register MFA. A password reset alone doesn't end a session the attacker already holds.
- Remove the attacker's footholds. Delete malicious inbox rules, forwarding settings and any app connections they added.
- Find out what they saw and sent. Review the mailbox audit log for messages read, searched and sent, especially anything sent to vendors or customers.
- Warn the people who might be targeted. Tell vendors and customers in the affected threads, by phone, not to act on recent payment changes.
- If money already moved, call the bank immediately and file a complaint at ic3.gov. Ask the bank to initiate a recall.
You can see how Elevaire handles this sequence in practice in what happens when we detect a threat.
How Elevaire Managed Security handles invoice fraud
Elevaire Managed Security covers the email and identity side of this problem alongside whatever IT provider or internal team a company already has. Your IT provider keeps running laptops, networks and day-to-day support. Elevaire runs the security layer on top: email security, multi-factor sign-in and access hardening, phishing tests and security training, and a quarterly security review.
On the Plus and Complete packages, that includes 24/7 detection and response and identity monitoring, which is where suspicious sign-ins and new forwarding rules get caught. The Essentials package provides business-hours response.
Example engagement: a 45-person engineering firm's finance manager reports that a supplier asked to update its bank details. Elevaire would confirm the request through a call-back to the supplier, then review the firm's own mailboxes for signs the request started internally. Elevaire would then tighten MFA on the finance team's accounts, block external auto-forwarding, add rules that alert on new inbox rules, and write a one-page payment change procedure the finance team can follow without calling anyone first.
Frequently Asked Questions
How do I know if our email account has been compromised?
Common signs include inbox rules nobody remembers creating, emails going missing from vendor threads, sign-in alerts from unfamiliar locations, and contacts replying to messages you never sent. Check your mailbox rules and forwarding settings first, since attackers rely on them to stay hidden. If you find something unexpected, treat it as a live incident and revoke active sessions, not just the password.
Does multi-factor authentication stop invoice fraud?
It stops most password-only attacks, which is why it's essential. It doesn't stop every one: some phishing kits capture the session after MFA is completed, and SMS codes can be intercepted. Phishing-resistant methods such as security keys or passkeys close most of that gap, and payment verification by phone covers the rest.
Who is liable if we pay a fraudulent invoice from a vendor's hacked email?
It depends on your contracts, your bank agreement, your insurance policy and, in some cases, the courts. A written call-back procedure for bank detail changes is one of the strongest positions you can be in, both for prevention and for any later dispute.
What does managed email security cost for a company our size?
Elevaire Managed Security uses predictable monthly pricing based on package and team size, with no hourly billing. For context on what's at stake, the FBI's 2024 figures put the average reported BEC loss at roughly $129,000 per complaint. Our security cost calculator gives an estimate by package and headcount.
How does this work alongside our existing IT provider?
Elevaire works alongside your current IT provider or internal IT team rather than replacing them. Your provider keeps handling devices, networks and support tickets, while Elevaire manages the security tools, watches for threats like compromised mailboxes, and coordinates any changes with your IT team.
What should we do first to reduce invoice fraud risk?
Start with the process control: require a phone call to a known number before any vendor bank detail change is paid. Then confirm MFA is on for every mailbox, especially finance and leadership, and that automatic forwarding to outside addresses is blocked.
About Elevaire Systems
Elevaire Systems is a managed security services provider for growing companies. We run your security tools and vendors, watch for threats and stop attacks for one predictable monthly price, with virtual CISO leadership and compliance readiness when you need more.
See Where Your Security Stands
Take the free 3-minute scorecard for your top security risks, or book a 30-minute Risk Snapshot to talk it through.

