July 10, 2026

Ring, ring. Fraudsters Are Calling Your Call Centers and Stealing Money

Synthetic voice attacks against insurance companies jumped 475% in 2024.*

That’s the highest sector-specific spike recorded anywhere in financial services, in a channel insurers can’t afford to lock down: 62% of insurance customers say they’d rather call than use any other channel.* Fraud networks have noticed the same thing policyholders have. Insurance call center fraud is now the fastest-growing way in and it doesn’t look like an attack. It sounds like a phone call.

For years, insurers have focused on securing their digital premises. Login pages required 2FA (two-factor authentication); and device, behavioral, and transaction monitoring became standard parts of the fraud stack.

Fraudsters have adapted and started targeting the call center. Unlike a login page, the call center was built to help people through exceptions: a locked account, a missing document, a payout question, a beneficiary change, a policy surrender. It was designed around empathy, speed, and resolution. Those are good things, but they’re also exactly what fraudsters exploit.

The call center became a fraud shortcut

Most agents are doing exactly what they were trained to do: help the person on the other end of the line. They’re not trained to hear the difference between a legitimate policyholder under stress and a trained fraudster using stolen personal information and AI-generated speech.

Fraud networks are attacking this blind spot at scale, and call center fraud has hit its highest level in six years. Account takeover fraud rose 37% from 2024 to 2025, and nearly two-thirds of financial institutions say most account takeovers start with a phone call, not a login.

The mechanics are simple and cheap and have a name: vishing, or voice phishing. A fraudster buys breached personal information. They call in with enough real details to sound credible. They may spoof a number, use a virtual line, or layer in a synthetic voice. Then they push for the action that pays: change the bank account on file, redirect a payout, update contact details, move a beneficiary, or surrender a policy.

By the time the suspicious claim or transfer gets reviewed, the most important decision has already been made. The caller was treated as the policyholder

Why the old checks are failing

A mother’s maiden name or the last four of a Social Security number is not a secret. Those details live in breached databases, brokered data sets, and fraud playbooks. Pindrop’s public summary of its 2025 report says 53% of fraudsters who attempted account takeover successfully passed knowledge-based authentication.

One-time codes help catch some fraud, but roughly one in four fraudsters passed OTP challenges. A code does not always confirm that the person is the rightful policyholder.

Once they pass the call center’s checks, everything downstream starts to trust the interaction. The account looks real because it is real, the policy has history, and the caller knows the right answers. The agent has no obvious reason to stop the request.

That is why insurance call center fraud is so hard to catch late. The fraud does not always look like blatant intrusion.

How do you fix this?

A modern call center defense should start before the conversation does. Verify and assess the risk of the call itself before the fraudster ever gets to a human agent.

Evaluate the call the same way a strong onboarding defense evaluates a new application, in real time and across multiple signals at once:

• Caller and device intelligence: Does the phone number, device, and carrier history attached to this call actually match the policyholder on file, or does it just resemble them?
• Behavioral and session risk: Does this call pattern fit how this policyholder normally interacts with your business, or does it look like a script a fraud network runs at scale?
• Cross-channel and cross-carrier signals: Is this same identity showing up in risk-flagged sessions elsewhere, a sign of a fraud network working a list rather than a single bad actor?

When those signals are evaluated together the moment the call comes in, the decision no longer depends on whether an agent asks the right follow-up question. The risk gets scored before a human ever has to guess.

Risky callers should get stepped up, routed, or restricted before they get the chance to change sensitive account details. Fraud teams should see the patterns across channels before any access is provided.

Call centers are now part of the identity perimeter, and insurers that understand that this is an enterprise identity issue instead of a call center issue can stop more fraudsters at scale.

A voice check alone can confirm a caller sounds human. It can’t tell you whether that same identity just tried three other carriers this week. That’s a different question and it’s the one that actually stops a fraud ring, not just a single call.

Agents get to focus on the quality of the interaction. Policyholders get an experience free of intrusive questions. And the fraud networks lose the blind spot they have been counting on.

Hang-up on fraudsters before they ever get to your human agents. Talk to an expert to see how continuous identity intelligence scores every call before it reaches an agent.

Sources

* Pindrop 2025 Voice Intelligence and Security Report ** | PDF Report_**

TransUnion H1 2026 Top Fraud Trends Report

TransUnion 2023 State of Omnichannel Authentication Survey

Dalia Woroniec

At Socure, Dalia is the Vertical Marketing Lead for Core Financial Services, focused on insurance, wealth management, lending and credit, and brokerage. She is a 20-year product marketing vet whose experience spans hyper-growth startups to big tech across fintech, travel, and ShaS, at companies like Google, AdMob, and Expedia. That range of industries and growth stages is exactly what she brings to the table, a cross-sector lens that sharpens positioning, accelerates market entry, and builds pipeline with intention.

Frequently Asked Questions

What is call center fraud in insurance?

Call center fraud happens when a fraudster impersonates a real policyholder over the phone to make changes the policyholder never approved, like redirecting a payout or updating a beneficiary. The account and its history are real; only the caller is not.

Why do fraudsters target call centers instead of logins?

Login pages are now protected by 2FA, device checks, and behavioral monitoring. Call centers were built around empathy and speed, not verification, so fraudsters exploit the same qualities that make good service good.

Can knowledge-based authentication (KBA) stop this?

Not on its own. In 2025, 53% of fraudsters attempting account takeover passed KBA, because the “secret” details it relies on are already circulating in breached data.

Does a one-time passcode (OTP) solve the problem?

OTPs help but aren’t sufficient alone. A code confirms someone has access to a phone, not that they’re the rightful policyholder.

What are red flags of a fraudulent call to an insurance call center?

A request that doesn’t match the caller’s history, a device or carrier profile that doesn’t match the policyholder on file, and the same identity appearing in risk-flagged sessions at other carriers.

How is call center fraud different from account takeover fraud?

Call center fraud describes the entry point: the phone channel. Account takeover describes the outcome: unauthorized access. Most account takeovers in insurance now start with a phone call, not a login.