That FaceTime From Your Kid? In 2026, There's a 1-in-3 Chance It's a Scam.

Here is the number that should stop you mid-scroll: nearly 500%. That is how much deepfake identity fraud is forecast to increase in 2026 compared to last year. Not double. Not triple. Nearly six times as much. And the scariest part isn't the math — it's what the math means. A face on a video call, a voice on the phone, a photo on a profile: none of those things are automatic proof of identity anymore.

Think about the last time you got a video call from someone you knew. You probably didn't question it for a second. You saw the face, you heard the voice, and your brain said: yep, that's them. That reflex — that instant recognition — is exactly what fraudsters are learning to exploit at industrial scale.

---

The Number Nobody Is Talking About Loudly Enough

The forecast comes from ASIS International, citing the Shufti Identity Fraud Index, and it lands like a gut punch: deepfake identity fraud is projected to increase nearly 495% in 2026. That figure isn't measuring deepfakes in general — political videos, celebrity hoaxes, memes. This is specifically about fake identities being used to steal, scam, and deceive real people out of money, access, and private information.

But here is where it gets genuinely unsettling. Research published in Scientific Reports found that humans correctly identify an AI-generated voice only about 60% of the time. Flip that around: you get it wrong four times out of ten. That's barely better than a coin flip. And that's for audio alone — add a convincing video face and your brain starts filling in the gaps, making you even more confident and even more wrong.

This is what psychologists call the availability heuristic — your brain judges how likely something is based on how easy it is to picture. A video call with your sister's face feels real because your brain has seen that face a thousand times. It doesn't stop to think: but is this ACTUALLY her? Fraudsters know this. They're betting their whole operation on it. This article is part of a series — start with Deepfake As A Service Fake Boss Scams Workplace Risk.

---

This Isn't Expensive Tech Anymore. That's the Problem.

A few years ago, making a convincing deepfake took hours of computing time, technical skill, and a lot of source footage. Today? In many cases, all it takes is a single photo and a text prompt. The barrier to creating a fake-but-convincing identity has collapsed — and the fraud industry has noticed.

According to data compiled by Bright Defense, 85% of organizations have already experienced a deepfake-related incident. That's not a future problem. That's a right-now problem. Meanwhile, Fintech Global reports that 42.5% of fraud attempts in the financial sector are now AI-driven — but only 22% of financial institutions have put AI-based fraud defenses in place. That is a very wide gap between the attack and the defense.

And it's not just video calls. Document deepfakes — think fake driver's licenses, fake passports, fake pay stubs submitted to verify an identity online — made up about 12% of deepfake fraud in 2025. Per the ASIS report, that category alone is expected to grow by 3,892% in 2026. That is not a typo. Nearly a 40-fold increase in fake document fraud. The attack surface isn't narrowing; it's widening in every direction.

"The most effective deepfake identity fraud attacks combine different techniques, such as blending a presentation attack with a synthetic identity or using an injection attack to skip the use of a camera to pipe an AI-generated synthetic identity straight into a verification tool." — Shufti Identity Fraud Index, as reported by ASIS International

Translation: the most dangerous scams aren't using just one fake thing. They're layering fake faces, fake documents, and fake identities all at once. It's not a con artist with a mask — it's an entire fake person, assembled from AI parts, designed to pass every check you throw at it.

---

What This Actually Looks Like in Real Life

Consider what happened to the engineering firm Arup. Employees were pulled into a video call with what appeared to be their CFO and several colleagues. The people on screen looked right and sounded right. The employees transferred $25.6 million before anyone realized none of those people had actually been on the call. Every face was AI-generated, according to StingRay's 2026 deepfake statistics analysis.

That's a company with trained finance professionals. Now imagine the same trick aimed at a parent whose "kid" calls in a panic needing emergency money wired. Or an elderly person whose "bank manager" video-calls to verify an account. Or an employee whose "CEO" sends an urgent request over a video message. These scenarios aren't hypothetical anymore. They're operational, scaled, and getting cheaper by the month. Previously in this series: Your Boss Just Called It Wasnt Him And It Cost 25 Million.

The fraudsters aren't even always inserting themselves into a live call. Some of the most advanced attacks now bypass the camera entirely — piping a pre-generated AI face directly into a video verification system so that a human reviewer or an automated check sees a convincing person who was never actually there. It's like a deepfake that happens before you even press "record."

---

So What Do You Actually Do About It?

Here's the honest answer: you can't out-stare a deepfake. Trying to spot one with your eyes alone puts you in coin-flip territory, and the technology is improving faster than human perception ever will. What you can do is change what you treat as proof.

The new mental model is this: trust arrives in layers, not in a single face. Before you send money, share something private, or grant access to anything important — even if the person on screen looks and sounds exactly right — ask yourself three things. Does this request fit the normal pattern of how this person contacts me? Is this coming through a channel we've used before? And does the reason for the request make sense without being weirdly urgent?

Urgency is a red flag, full stop. Deepfake scams almost always come with pressure — "I need this now," "don't tell anyone," "it's an emergency." Real people who actually know you don't usually demand instant action with no room for a quick callback. If someone is pushing you to decide in the next five minutes, that pressure is the tell — not the face.

Build yourself what security professionals call a "second channel" — a way to verify that doesn't depend on the conversation you're already in. Call back on a number you already have saved, not one they give you. Send a message on a different app. Ask a question only the real person would know. It sounds fussy until you realize that a $25 million engineering firm got cleaned out because nobody paused to do exactly that. Up next: Your Boss Just Called It Wasnt Him And It Cost 25 Million.

If you've ever looked at a photo, a profile, or a video and wondered whether the person in it is actually who they claim to be — that instinct is right and it's worth acting on. Tools that let you compare a face against other verified images of the same person (checking whether the face matches across multiple real-world sources, rather than just taking a video at face value) are exactly the kind of second-layer check that defeats a deepfake's best trick. The face might be perfect. The match won't be.

According to Keepnet Labs, deepfake-related fraud incidents have already seen staggering surges — with some voice-cloning categories spiking over 1,300% in recent reporting periods. The 2026 forecast isn't an alarm being raised for the future. It's a description of something already in motion.

And Thomson Reuters Institute notes that even courts are grappling with this: judges now need specific guidance on how to evaluate whether video or audio evidence submitted in legal cases is genuine. If the legal system is scrambling to catch up, the rest of us are definitely not overreacting by being more careful.

---

Here is the question worth sitting with tonight: if someone you love called you right now — face on screen, voice exactly right, maybe even a little emotional — and asked for money or access to something private, what would make you 100% certain it was really them? Because "they looked like themselves" is no longer the answer. And the people building these scams are counting on the fact that most of us haven't figured out our answer yet.