Deepfake scams: Singapore acts as fraud attempts jump 1,300%
Quick answer
How do you protect yourself from deepfake video call scams?
Confirm identity through a second channel that does not rely on a face or voice, such as calling back on a known number. Researchers found people spot a high-quality deepfake only about a quarter of the time, so a convincing video call is not reliable evidence. Liveness detection can add a further check.
Deepfake scams: Singapore acts as fraud attempts jump 1,300%
Watch This Episode
This episode is based on our article:
Read the full article →Deepfake scams: Singapore acts as fraud attempts jump 1,300%
Full Episode Transcript
Picture this. You join a video call with your company's chief financial officer. You can see his face. You can hear his voice. He tells you to wire millions of dollars, urgently. So you do it. And every single person on that call was fake. Not one real human, except you.
That's not a hypothetical
That's not a hypothetical. Singapore's police reported it as one of the most convincing A.I. impersonation cases they've ever seen. And if you've ever been on a video call, a work meeting, a family chat, anything, this story is about you. Because the thing we all rely on to know someone is real, their face, their voice, that proof just stopped working. Singapore's financial regulators spent this September admitting exactly that. Fraud attempts using deepfakes jumped more than thirteen hundred percent in a single year. So how do you protect yourself when your own eyes and ears are lying to you?
Let's start with that number, because it's staggering. According to fraud researchers, deepfake scam attempts went from about one a month to roughly seven a day. In one year. That's not gradual growth. That's an explosion. And in that Singapore boardroom scam, the attackers didn't fake one person. They generated several fake identities, bosses, colleagues, all rendered live, in real time, moving and talking like a normal meeting. One reported case cost the victim around five million Singapore dollars. That's close to four million U.S. dollars, gone.
Now here's the part that should stop you. Researchers found that when people watch a high-quality deepfake video, they spot the fake only about a quarter of the time. Think about that. Three out of four times, your brain says "that's a real person", and it's wrong. For the rest of us, that means the confidence you feel watching a video call? It's no longer reliable evidence of anything.
You'd expect banks and companies to be racing to
So you'd expect banks and companies to be racing to fix this. They're not. Only about one in five financial institutions have put A.I.-based defenses in place. Meanwhile, researchers say more than four in ten fraud attempts hitting the financial sector are now A.I.-driven. One side is speeding up. The other is barely moving. That gap is where the money disappears.
Singapore's response is worth understanding, because it's practical. They pulled small levers, not just big ones. For example, regulators removed nicknames from PayNow, that's their instant payment system. Why? Because a friendly display name made it easier for a scammer to look like your bank or your friend. They also pushed banks toward what's called liveness detection. In plain terms, the system checks that a real, three-dimensional person is present. It watches for real motion, real skin texture, and sometimes asks you to turn your head or perform an action. A flat video replay can't do that convincingly.
And those steps worked. Singapore's earlier safeguards, rolled out the year before, actually brought impersonation scam cases and losses down. So this isn't hopeless. It's just unevenly applied.
The Bottom Line
But here's the uncomfortable truth Singapore's regulators landed on. You cannot fix this with technology alone. Because in many of these scams, the victim authorizes the payment themselves. The system did its job. The person got fooled. Which means the last line of defense isn't a better camera, it's a second way to confirm someone is who they say they are, one that never depends on a face or a voice.
So let's bring this home. Deepfakes have made fake faces and fake voices good enough to fool almost anyone, almost every time. Scam attempts are exploding, but most institutions haven't caught up. And the real fix isn't just smarter software, it's never trusting a single video or call to prove who someone is. Whether you approve wire transfers for a living or just answer a call from someone claiming to be your bank, the rule is the same now. Verify through a channel they can't fake. The full story's in the description if you want the deep dive.
Ready for forensic-grade facial comparison?
Full forensic reports with detailed similarity scoring. Results in seconds.
Run My First SearchMore Episodes
AI Voice Cloning: Fake Call Takes Rs 1.2 Lakh From Florist
A florist in Kolkata picked up the phone and heard a voice she'd known for years. It was a friend. It sounded exactly like her. It wasn't. By the end of the night, the florist had lost one point two lak
PodcastTikTok's Response: Nearly 50 Nude Deepfake Apps, 'Not Ads'
Nearly fifty apps that can turn an ordinary photo into a fake nude were reportedly being promoted on TikTok. And when researchers raised the alarm, TikTok's answer was simple. Those weren't ads. <break
PodcastFacial Recognition in Stores: False Match Ejects UK Shopper
A couple walked into a Budgens grocery store in the U.K., and a staff member told the wife she was accused of stealing "loads of meat." Both of them say they hadn't been in that store in months. They we
