Sentinel Underwrite OpsEnterprise Underwriting Intelligence
Claims & Coverage

Crime Insurance vs Cyber Insurance: Which One Pays When Money Leaves the Account?

A spoofed invoice, a hijacked bank login and a dishonest bookkeeper each trigger different insuring agreements. Five payment-loss scenarios, which policy pays for each, and which one to buy first.

By the Sentinel Underwrite Ops editorial team · November 2026 · 8 min read

The Coverage Question to Settle Before the Next Wire Goes Out

When money leaves your account and lands with a fraudster, your broker's first question is what actually happened. The answer decides whether your crime policy, your cyber policy, both or neither responds. Most buyers frame this as crime insurance vs cyber insurance, as if one replaces the other. It doesn't. The two are built around different triggers, and the gaps between them are where payment-fraud claims usually get denied.

This piece runs five common payment-loss scenarios through typical policy wordings: funds transfer fraud, computer fraud, social engineering, insider theft and lost client funds. It then says which policy to lead with for different kinds of companies. Wordings vary widely by insurer and country, so treat this as a map for your broker conversation, not a coverage opinion.

Advertisement

Why Crime and Cyber Wordings Respond to Different Events

A crime policy insures direct loss of money, securities or property. It's sold as commercial crime in North America, as crime or fidelity in the UK and Europe, and often as fidelity guarantee in India, the Gulf and parts of Asia and Africa. It's first-party, often written on a discovery basis, and split into separate insuring agreements: employee theft, forgery, computer fraud, funds transfer fraud and, by endorsement, social engineering.

A cyber policy is built around network security and privacy events: breach response, forensics, notification, business interruption, extortion and third-party liability. Stolen money was an afterthought. Many cyber forms exclude loss of funds outright or give it back through a "cyber crime" extension with a small sublimit.

Two consequences follow. Crime policies care about how the money moved and who was deceived. Cyber policies care about whether your systems were compromised. A loss can pass one test and fail the other.

Scenario One: A Fraudster Instructs Your Bank Directly

Say an attacker phishes your treasury manager's bank portal credentials, logs in and sends $180,000 to an overseas account. Nobody at your company approved it. The bank acted on instructions that appeared to come from you.

This is the core case for funds transfer fraud coverage under a crime policy, which responds to fraudulent instructions sent to a financial institution without your knowledge or consent. It's typically covered up to the crime limit, less the deductible.

A cyber policy may respond too, if it carries a funds transfer or cyber crime extension, but expect a sublimit, often well below the crime limit on small and mid-market forms. Cyber would also pay forensics to establish how the credentials were taken, which the crime policy generally won't. Lead policy: crime. Cyber covers the investigation.

Scenario Two: An Intruder Changes Payment Data Inside Your Systems

Now the attacker gets into your ERP or accounts-payable platform and quietly edits a supplier's bank details. Your next payment run sends $95,000 to the attacker. Your staff released the batch, but didn't know the data had been altered.

Computer fraud agreements typically cover loss resulting directly from fraudulent entry or change of data in your system. That word causes most of the disputes. Insurers sometimes argue a human approval step breaks the causal chain, and courts in different jurisdictions have split on similar facts. Ask whether the wording says "resulting directly from" or something broader, and whether a payment run approved in good faith is carved back in.

The computer fraud vs social engineering line matters here. If the fraudster changed the data, it's computer fraud. If the fraudster emailed your clerk and the clerk changed it, many insurers will call it social engineering, which carries a much lower limit. Lead policy: crime, with cyber covering forensics and any data breach costs.

Scenario Three: An Employee Is Tricked Into Paying

This is the most frequent loss: business email compromise. A message that appears to come from your CFO, or from a real supplier's spoofed or hijacked mailbox, asks your AP clerk to wire $60,000 to new bank details. The clerk does it.

Standard computer fraud and funds transfer fraud agreements usually exclude this, because an authorized employee voluntarily made the transfer. Cover comes only from a social engineering endorsement on the crime policy or a social engineering sublimit on the cyber policy. Sublimits are commonly in the tens to low hundreds of thousands of dollars, and specialist crime markets may offer more if your controls hold up.

Read the conditions. Many endorsements require that you verified the change of bank details by calling a number already on file, not one taken from the email. Skip that step and the claim can fail even though the endorsement exists. Lead policy: whichever offers the higher social engineering limit with conditions your AP team actually follows. For most companies that's the crime policy.

Scenario Four: The Thief Is on Your Payroll

Say a finance employee in a regional office adds a fictitious vendor and pays it $12,000 a month for two years. Total loss: $288,000, found at year-end audit.

This is employee theft coverage, the original purpose of fidelity insurance. Crime policies cover theft by employees, often including temporary staff and, by extension, contractors. Discovery-based forms can respond to losses that ran for years before you found them, subject to any retroactive date.

On fidelity insurance vs cyber, the answer is one-sided. Cyber policies almost always exclude dishonest acts by your own employees aimed at stealing money. Some cover a rogue employee who leaks data, but not a payroll embezzler. Lead policy: crime, every time.

Scenario Five: Your Client's Money Goes Missing

Two versions keep coming up. In the first, a law firm, conveyancer, payroll bureau or property manager holds client funds and a fraudster diverts a $400,000 completion payment from the client account. In the second, a criminal impersonates you to your customer, and the customer pays your invoice into the criminal's account.

For the first, check whether your crime policy covers property of others in your care, custody or control, via a client property or third-party extension. Professional indemnity may respond if the client sues for negligence. Regulators in jurisdictions such as the UK, Australia, Singapore and the EU generally expect firms holding client money to show a clear source of reimbursement.

For the second, the money was never yours, so crime policies generally won't pay. Some cyber policies offer an "invoice manipulation" extension that covers your uncollected receivable when your compromised system was used to deceive the customer. Lead policy: crime with client property cover for fund holders; cyber with invoice manipulation for everyone else.

Scenario Matrix, Price Ranges and Red Flags

ScenarioCrime policyCyber policy
Funds transfer fraudCore cover, full limitExtension, often sublimited
Computer fraudCore cover; check "directly" wordingExtension plus forensics
Social engineeringEndorsement, sublimited, callback conditionsSublimit on many forms
Insider theftCore coverGenerally excluded
Client funds / invoice redirectionClient property extensionInvoice manipulation extension

On cost, these are indicative 2026 market ranges that vary by contract, country and loss history, so confirm them with your broker. A standalone crime policy with a $1 million limit for a company of 50 to 250 staff might run roughly $2,000 to $10,000 a year. Raising the social engineering sublimit usually adds a meaningful percentage to premium. A $1 million cyber policy for a similar firm might cost roughly $3,000 to $20,000, depending on industry and controls. Revenue, finance headcount, number of payment approvers, dual-authorization controls and prior BEC losses drive both prices.

Red flags in the wording:

  • A social engineering sublimit below your largest routine supplier payment
  • Callback conditions that don't match how your AP team really verifies changes
  • "Other insurance" clauses in both policies, each pointing to the other
  • Employee theft definitions that exclude contractors or offshore shared-service staff
  • Territorial limits that exclude subsidiaries where payments are processed

Verdict: Which Policy to Lead With, by Buyer Profile

Firms under about 50 staff with outsourced IT: lead with cyber, but only if the form includes funds transfer, social engineering and invoice manipulation extensions with meaningful limits, ideally at least $100,000 each. Breach response matters more to you than a big crime limit, and bundled forms are the efficient buy.

Mid-size companies with an in-house AP team and hundreds of vendors: buy both. Lead with crime for payment losses, sized to your largest realistic single transfer, and treat cyber's crime extensions as backup. This is where the crime insurance vs cyber insurance gap does the most damage, because one diverted supplier payment can exceed every cyber sublimit.

Large and multinational groups: build a dedicated crime program with local policies or a compliant global master where non-admitted cover is restricted, as it is in several markets including parts of Latin America, India and China. Keep cyber focused on breach, extortion and business interruption.

Firms holding client money: crime with client property cover is non-negotiable. Pair it with professional indemnity and confirm what your regulator requires.

Next step: pull your three largest payments from the past year and ask your broker, in writing, which insuring agreement and limit would respond if each had been diverted by email, by system intrusion and by an insider. Any answer of "sublimit" or "excluded" tells you which policy to fix first. A licensed broker or coverage counsel should confirm the final structure.

Frequently asked questions

Does cyber insurance cover money stolen through business email compromise?

Only if the policy includes a social engineering or cyber crime extension, and those are usually sublimited, often well below the main limit. Base cyber forms generally exclude loss of funds. Compare the sublimit with your typical payment size.

What's the difference between computer fraud and social engineering coverage?

Computer fraud covers losses caused directly by someone fraudulently entering or changing data in your systems. Social engineering covers losses where an authorized employee is deceived into making the payment voluntarily. Insurers draw this line carefully because social engineering limits are usually much lower.

Is fidelity insurance the same as crime insurance?

Largely, yes. "Fidelity" or "fidelity guarantee" is common in the UK, Europe, India and the Middle East, while "commercial crime" is the usual term in North America. Modern forms typically bundle employee theft with computer fraud and funds transfer fraud, but scope varies, so compare the insuring agreements and not just the policy title.

Will both policies pay if a loss triggers crime and cyber cover?

You can't recover the same dollar twice. When both policies respond, "other insurance" clauses decide which pays first, and if each points to the other, settlement can drag. Ask your broker to coordinate the wordings so one policy is clearly primary for payment losses.

What controls do insurers expect before offering higher social engineering limits?

Most want documented callback verification for any change of bank details, using a number on file. They also commonly expect dual authorization above a set threshold, multi-factor authentication on email and banking portals, and staff training on payment fraud. Demonstrating these controls usually helps with both limits and pricing.

Written by the Sentinel Underwrite Ops editorial team. Prices, features and rules change often - confirm details with the vendor or a licensed professional before you decide.