We hear a lot about AI. It’s practically everywhere. In advertisements, in the news, in the media, on our phones and computers.
But what exactly is AI?
There are many definitions, and even experts often disagree on what AI truly is. A widely accepted general description is this:
“Artificial Intelligence (AI) refers to machines’ ability to perform tasks that require human-like intelligence. These include reasoning, learning, planning, decision-making, problem-solving, perception, communication, and creativity.”
The simplest and most widely known form of AI isn’t actually true artificial intelligence. According to the official terminology, these are called “generative AI.” This means that a neural network-based algorithm is fed with an enormous amount of data by developers, and when it receives a request, it tries to “guess” the most appropriate answer based on all that data.
The response can be a simple text, but also a generated image. These are the solutions we encounter most often. Among the text-based tools, ChatGPT was the first globally well-known program that became mainstream.
These are the programs we can talk to, that draw whatever we ask for, answer our questions, create music for us, or even turn a single photo of us into a nearly lifelike video.
Basically, these algorithms are programmed for “positive thinking.”
They won’t give us ethically questionable answers. They don’t use profanity, don’t respond to racist questions, don’t discuss sexuality, and don’t generate sexual or violent content—neither in text nor in image. Of course, this applies only to open systems. For private purposes, unfortunately, it’s possible to set up versions without such restrictions.
But it’s important to understand that these programs do not think!
They don’t “understand” what we’re asking—they only interpret the input and, based on the massive amounts of data they’ve been trained on, try to give us the most relevant answer.
If we ask such a program to write a letter on a given topic, it first interprets the request, and then, without asking anything in return, gives us its best attempt. If we ask in Hungarian, it answers in Hungarian; if we ask in English, it answers in English; if in French, then French. The structures of different languages are so well integrated that it can respond in nearly perfect grammar in almost any language.
And this is exactly the point—where things can go very wrong!
This is the point where language barriers disappear. The point where, at nearly zero cost, someone anywhere in the world can use a generative AI tool to communicate in the language of their intended victim. It could be a simple email—but now, it can also be real, audible speech! These AI solutions can now generate not just text or images, but video and human-sounding voice, too.
With their help, perfect human profiles can be created—complete with photos, vacation videos, personal stories. Of course, this goes beyond free tools like ChatGPT—you’ll need a bit more knowledge, but it’s knowledge that can be acquired in just a few weeks.
It works so well that hundreds of fake accounts are created every day on Instagram—accounts that aren’t run by real people, but by AI-generated content.
Beautiful, flawless men or women. Who have everything. Who constantly travel the world. Who take selfies and post their lives everywhere they go. Who message you. Who start talking to you. Who get to know you. Map you. Learn your patterns.
And once they’ve earned your trust, the scam begins.
Maybe it starts with asking for money—and ends with full access to your bank account, draining it into an untraceable wallet that even the police can’t recover.
But it’s not the AI doing this to you.
It’s not the AI gaining your trust.
It’s not the AI stealing from you.
Behind every scam, there is always a human. A person who uses generative AI simply as a tool. Just like you or I—but instead of writing a thesis, they use it to scam you! Perfectly written emails. Flawless-sounding voice or video messages. Or maybe a heartbreaking story, ending with tears in a voice note asking you to send money because their little boy is sick and they can’t afford medicine.
It’s not AI you should be angry with.
And the developers who built these tools are not to blame either.
What we’re dealing with is simply crime.
Criminals who want to steal from you.
Criminals who use the latest technology against you.
Criminals who didn’t study for years to master this—just sat in front of a computer and, in a few weeks, learned enough to use generative AI to make you believe anything.
Maybe the person attacking you isn’t even of legal age.
Maybe they live on the other side of the world.
Or maybe… they’re sitting in the office next door.
AI can’t think.
It doesn’t know it’s deceiving you, hurting you, robbing you.
It simply executes the command… given by a criminal.
