That Familiar Face Promising You Money? Only 0.1% of Us Can Tell It's Fake

It's 11 p.m., you're half-asleep, scrolling before bed, and a video stops your thumb cold. It's a finance guy you've seen on the news a hundred times, calm and confident, explaining a crypto deal that sounds too good to skip. He sounds like himself. He looks like himself. He is not himself. He's a deepfake — a fake video made by AI software that borrows a real person's face and voice and makes them say things they never said. And that exact trick just helped Australian officials rip down more than 3,100 fake crypto trading sites in a single year.

The Big Number That Isn't Actually Good News

Here's the thing nobody says out loud when a regulator announces a big takedown: the number is only impressive because the problem got so much worse. Australia's securities watchdog (the government agency that's supposed to keep investment offers honest) pulled down 3,106 crypto platforms and more than 19,400 scams total during its last fiscal year, according to Finance Feeds. That's a 30% jump in crypto scams alone, and total scam removals surged 182% year over year.

Read that again. Not 30% more scams caught because enforcement got better. Thirty percent more scams because there are simply more of them to catch, and AI tools are why. A criminal used to need writers, designers, and someone who could fake a British or American accent convincingly. Now they need a laptop and a few dollars of AI credit.

How the Fake Video Factory Actually Works

These scams don't show up alone. That's the part people miss. A fake investment video rarely stands by itself — it's stitched into a whole web of fake supporting evidence: a fabricated news article about the "opportunity," a spoofed version of a real news site, fake glowing reviews, and sometimes even a stolen government licence number (in Australia, legitimate investment firms carry something called an AFSL — basically a permission slip proving they're allowed to sell investments legally). Scammers lift real firms' licence numbers and slap them onto fake products, so even checking the paperwork doesn't always save you.

Each fake piece props up the others. You see the video, then Google it, then find the fake news article "confirming" it, then see comments praising it — and every layer feels like independent proof, when really it's one liar wearing five different masks.

That's not a typo. Basically none of us can trust our own eyes and ears anymore when it comes to a screen. And the money numbers back up how badly that's already gone wrong: Australians reported A$837.7 million in investment scam losses in 2025. A slice of that — A$7.4 million — came specifically from cases where the victim actually noticed a deepfake was involved. But that number is almost certainly way too small, because most people who get fooled by a fake video never realize it was fake, and plenty never report it at all.

"AI is making investment scams more convincing and harder to detect. A simple online search is not enough to verify whether an opportunity is legitimate. The presence of polished content, familiar branding or convincing testimonials does not mean an investment is genuine." — ASIC Chair, as reported by GNCrypto News

Sit with that for a second. The head of the agency whose entire job is catching this stuff is saying: even our own checks aren't enough anymore. That's not regulator caution talk. That's an admission that the old rules for spotting a scam — "does it look professional, does it show up in a search, does someone I trust vouch for it" — are broken.

Why This Is Spreading So Fast

The scale of this jump isn't small-time. Fraud tracked to AI tools and deepfake videos in the crypto world has climbed 13 times over compared to 2022, according to Bitcoin World. Separately, deepfakes built around Australian political figures have been tied to at least A$5.3 million in losses, per Bitcoin News. And in the U.S., the FBI logged $632 million in AI-related investment fraud losses in 2025 alone, with projections showing deepfake case counts climbing nearly 3,900%, according to Memeburn.

This isn't a bug in the system. It's the system working exactly as designed — for criminals. A phishing setup that used to take a scammer roughly 16 hours to build by hand can now be assembled in under five minutes with AI tools doing the writing, the voice, and the video. Which means one person, alone, at their kitchen table, can now run what used to take an entire criminal crew. Continue reading: That Familiar Face Promising You Money Only 0 1 Of Us Can Te.

The One Thing You Can Actually Do

If you've ever paused a video mid-scroll and thought, "wait, is that actually them?" — trust that instinct. It's the right question. Here's the part that trips people up: they try to verify the video by watching more videos, reading more articles about it, checking more reviews. That's like fact-checking a rumor by asking the person who started it. It just feeds you more of the same fake web.

The move that actually works is going around the video entirely. Find the person or company through a channel you had access to before you ever saw that clip — their verified account, a number you already had saved, the company's official press line. If the deal is real, that person or company can confirm it through a door the deepfake never touched. If they can't, or if suddenly there's "another way to reach them" conveniently offered in the video's comments — that's your answer.

What's Actually Coming

Here's where this is headed, and fast: a feed where the video, the branding, the "expert" talking, and the glowing comments underneath are all manufactured by the same person in the same afternoon — and none of it tells you anything true. Losses tied to deepfake scams in just the first half of 2026 have already blown past the entire total for 2025 by 263%. That's not a slow climb. That's a curve going nearly straight up.

So next time a familiar face promises you money is waiting — ask yourself the boring, unglamorous question instead of the exciting one. Not "how do I get in on this?" but "would this actual person really pitch me over a random video ad?" Nine times out of ten, they wouldn't. The scammers are betting you're too tired, at 11 p.m., to stop and ask.