Blog
The Callback Rule: Why Carriers Ask How You Send Money
Some cyber-insurance questions are about technology. This one is about habits. In our review of 13 public carrier applications (November 2026), 7 asked, in almost the same words, whether you confirm a request to move money, or a request to change a vendor's bank details, by a second method before acting on it.
Usually that second method is a phone call to a number you already had on file, not a number in the email that made the request. It's called a callback, and it may be the single cheapest control on the whole application.
What the forms ask
The wording varies, but the idea is consistent:
- Coalition splits the question by amount, asking about transfers over $5,000 and over $25,000, and asks separately about changes to banking details.
- At-Bay asks whether two people must authorize wire transfers over $25,000.
- Hartford, Tokio Marine HCC, Markel and Corvus all describe the same callback idea: verify by phone, using known contact information, before money moves or bank details change.
The specific dollar thresholds come from the forms. Your own threshold is a business decision.
Why it's on a cyber form
This isn't a hacking problem in the movie sense. It's what's usually called business email compromise, or BEC: a criminal uses email to persuade someone in your company to send money to the wrong place.
It comes in a few familiar shapes:
- The changed bank details. A real vendor's email account is taken over, or a lookalike address is used, and a polite message says the vendor has a new bank. The next legitimate invoice gets paid to the criminal.
- The urgent executive. A message that appears to come from the CEO asks the controller to send a wire quickly and quietly, often while the CEO is known to be traveling.
- The redirected payroll. An employee's email asks HR to change their direct-deposit account.
The FBI's Internet Crime Complaint Center (IC3) tracks business email compromise as one of the costliest categories of reported internet crime year after year; its annual reports show the trend. These schemes work because they look like normal business, arrive at busy times, and come from someone the recipient trusts.
Why a phone call works
The criminal controls the email conversation, and may control the phone number written in the email too. What they can't easily control is the number already sitting in your vendor file from before the request arrived, or the CEO's cell phone that you already have.
A callback to a known number breaks the scheme because it moves the conversation to a channel the attacker doesn't own. It takes two minutes. It doesn't require software. And it works even when the fake email is flawless, which more and more of them are.
Where it breaks down
When I ran IT at a company of about 200 people, the weak point was never the policy on paper. It was the moment. Month-end, a backlog of invoices, a trusted bookkeeper who had "always just handled it," and a message that sounded exactly like the vendor. That's when steps get skipped.
A few patterns make the rule stick:
- It applies to everyone, including the owner. If the CEO can bypass it with a phone call that says "just send it," it isn't a rule. The CEO should be the first to say, out loud, "If a request looks like it's from me, call me back anyway."
- The number comes from your records, never from the request. Not from the email signature, the invoice letterhead, or a "please call me at this new number" line.
- Changes to bank details are treated as high-risk on their own, even when no payment is due yet. That's where most of these schemes begin.
- Two people for larger amounts. One person enters the payment and a second approves it. Most business banking platforms support this, and it's often just a setting.
- No penalty for slowing down. The person who delays a payment to verify it should be thanked, even when the request turns out to be real.
Ask your bank what it offers
Your business bank is an ally here. Many banks offer features such as dual approval for wires and new payees, alerts when payee details change, and positive pay (where the bank checks payments against a list you provide before releasing them). Ask your relationship manager what's available on your account and what's switched on.
Also ask what to do if a fraudulent wire goes out. Speed matters. The bank's fraud line and the IC3 complaint process both belong on your one-page incident plan.
The insurance angle
Many cyber policies cover this kind of loss only up to a smaller limit, often called a social-engineering sublimit, and some require that your verification procedure was actually followed. That's another reason the question is on the application. Read your policy for the specifics, and ask your broker how your coverage treats a payment someone was tricked into sending.
Put it in writing, briefly
You don't need a manual. A two-line rule that everyone in finance and HR has read is enough:
No new payee, and no change to anyone's bank details, without a call to a number already in our records. Payments over [your threshold] need a second approver.
Post it where payments are made, include it in onboarding for finance and HR staff, and review it once a year.
What to do Monday morning
- Walk the payables process with whoever releases wires and ACH payments. Ask what happens today when a vendor emails new bank details.
- Write the two-line rule: callback to a known number for new payees and bank changes, and a second approver above a threshold you choose. The forms often use $25,000 as a reference point.
- Apply the same rule to payroll direct-deposit changes in HR.
- Call your bank and ask which controls are available: dual approval, payee-change alerts and positive pay. Turn on what fits.
- Announce the rule the same day, and say plainly that it applies when the request appears to come from you.