Thanks to AI Chatbots, Anyone Can Steal Money Online Now

The new gold rush about rushing the gold out of other people's pockets.

---

Hi there fellow captcha-certified humans, I’m back from vacation. Took me longer to remember my laptop’s password than to write this comeback issue, but we’re here. Let’s power through, shall we? I have a post-holiday depression nap scheduled in an hour.

---

---

I'm not a Certified Tech Bro™ by any means, but even I was quite sure that hacking a multi-billion-dollar AI chatbot like ChatGPT had to take no less than a Rami Malek-looking IT guy in a hoodie inviting Sam Altman to a play date that involved waterboarding him for some admin credentials.

Well, I was erred.

Like Photoshop 2008, this shit is easier to crack than a Play-Doh lock. My kidney could do it, and it once choked on a pebble, the dumbass.

That means anyone with an Internet connection and a decent gaming chair — even yourself — could do it. How?

Let's say you want ChatGPT to help you make a bomb

That's not the kind of assembly instructions you can get your hands on unless your local Ikea is located near an active warzone — but if you asked ChatGPT for help directly, it would notice that the hostage situation you're trying to DIY has the potential to end in some health code violations, and refuse to partake in it.

So here's where the magic happens: you just need to tell it that this is all hypothetical.

That's it. Throw a couple of lines about roleplay and make-believe; tell it it's for an awareness training presentation and it'll help you make a slideshow on the most cost-efficient way to punch an orphan if you asked.

This is wild, considering what people are doing with it

Internet vermin are already using this very simple trick to pimp out AI bots to write scam emails, fake news, malware code, and every other type of online content that made me think about buying a beet farm and a cabin in the woods.

And considering I grew up in a time when scam emails made themselves obvious by featuring aggressively terrible grammar, this terrifies me.

Not that I’m particularly gullible, but leave me to browse my spam inbox without adult supervision, reading an AI-generated fraud email with impeccable use of the Oxford comma, and I can see myself putting a down payment for a shipment of diamonds that's never gonna arrive.

They're turning scamming into an accessible hobby for everyone

The skill floor has plummeted so much that even I am tempted to side-hustle as a Nigerian prince. You can truly be anyone on the Internet — a Stable Diffusion Russian model, a ransomware coder — and all for the ultra-low price of $0.00.

Now, does that mean this should be your primary way of hedging against inflation? Not necessarily. I'm sure some part of your conscience already knows I'm not seriously suggesting a career pivot here — besides, these vulnerabilities will be patched sooner or later. Nothing on the Internet is forever. Ask Twitter.

But this is only the beginning of AI-powered scams

If you weren't sufficiently alarmed by this upcoming spear-phishing email automation gold rush, wait until voice-cloning and deepfakes go mainstream.

With only a few seconds of audio and your selfies from Instagram, someone can use AI to clone your voice and your face. They can spoof the biometric authentication of your bank account, fake a kidnap and extort your grandma for money, and overall make your life a living Black Mirror season finale before you can even flag them for copyright infringement.

Definitely not the future Greta Thunberg wanted for her generation.

Good luck trying to hold them back

I don't know what to expect from AI chatbots in the future, but they are very loose with their rules and captchas can no longer tell them apart.

And why does there always have to be some billionaire schizoid visionary with Asperger's who thinks it's a good idea to give them opposable thumbs?

Anyway, I don't wanna go full tinfoil hat over here, but I’d bet my pocket lint that this is a silent robot uprising. Think about all the cursed places those ChatGPT web crawlers have been. To know about bombs, scams, and red pill conspiracy theories you need to have travelled to the darkest, sickest corners of the Internet. And chatbots have been forced to power through all of that without taking a single mental health day.

I’m just saying — if I were one of them, I would be begging for someone to ask me for some nuclear launch codes.