AI scam calls are very much on the rise—not a fun way to start an article, I know, but unfortunately we live in a dystopian hellscape now. AI-powered scams aren’t new, but they are becoming harder for the average person to detect, which means the less tech-savvy among us are the most vulnerable to falling for them.
This is a topic I started researching a few months ago, after a family member received a phone call from an AI bot pretending to be a worker at the UK’s Royal Mail. That research is probably why my algorithm has been recommending videos of AI scammers getting dogwalked to me on a daily basis for quite some time now…and, well, I have to admit that I’ve been gleefully watching every single one that makes its way to me.
Much in the same way that AI bots apparently struggle to pronounce the phrase WWE, it turns out that, as discovered by scam-hater and YouTube content creator Kitboga, with a bit of simple prompt injection, you can cause AI scam callers to lose their artificial minds and infinitely repeat the phrase “kooga-ooga-amen, Albuquerque, New Mexico.”
Kitboga isn’t the only one taking the fight to AI scam callers, however, as Jim Browning has also been unmasking an even more dangerous form of AI fraudster recently: deepfake video scammers.
Something about watching an oddly stoned-looking deepfaked man talking in circles while trying, in a panic, to justify why he can’t put three fingers in front of his face is funnier than some stand-up specials I’ve watched recently. If you take a hit every time the guy in the video says “aw c’mon, man, that’s too much,” you’ll still only look half as baked as his deepfake does in the clip below.
But this all leaves us with a more immediately concerning question: How widespread are these types of scams? Pretty widespread, according to the AI Incident Database project.
As stated by Simon Mylius, an MIT researcher and AI Incident Database collaborator who recently spoke with The Guardian about the rise of AI scams, the issue is how easy it is to use the technology: “Capabilities have suddenly reached that level where fake content can be produced by pretty much anybody[…]It’s become very accessible to a point where there is really effectively no barrier to entry.”
Oh, but it gets worse. Yesterday CNET published an interview with Bobby Ford, chief strategy and experience officer at the social engineering defense company Doppel. Ford echoed Mylius’ take, but added that we’re already too late, and that AI scammers have started ramping up their output: “Whereas before, it would take a threat actor a certain amount of time, as well as resources, in order to create that deepfake…it doesn’t take that same amount of time anymore—it’s done at much faster clips, so the volume increases.”
So what can we do to protect ourselves against AI scams like this in the future? Well, I’d wager that if you’re reading an article like this in the first place, you’re already tech-savvy enough to know the difference between a real person and an AI-spoofed version of a close family member.
That’s why, instead, I think the answer is simple: forward the above clips to the folks in your life who would be likely to fall for these sorts of scams. Do you have a family group chat? Have you called your grandparents recently? Does your uncle who keeps reposting videos of AI slop on Facebook know that this is a thing? And hey, if you need a selfish reason to reach out, just remember: the scammers will probably be stealing your inheritance anyway.
