News JVTech Here’s how cybercriminals are already using ChatGPT to scam you
Criminals have not hesitated to rush into the breach opened up by artificial intelligence with text prediction. We thus find ChatGPT which can be used to create real fraudulent emails and much more.
ChatGPT can write everything and that’s the problem
The text prediction tool capable of writing text with a lot of humility or even lines of code can make it possible to create a scam from scratch and as an amateur. The ease of use means that anyone can get started. The basic skills needed are really limited.
The website “
check point research
” thus reveals that on certain forums hackers are already using ChatGPT to set up scams from scratch. This new threat, however, starts from the opening to the general public of the artificial intelligence of Open AI.
An example of content that one could qualify as “grey”, because it is not illegal, but little moral is the sale of chapters written by ChatGPT. We can thus find electronic books that have not had any work from a human.
ChatGPT scam in a nutshell
Check point research has also discovered in discussion threads on hacker forums the possibility of creating an algorithm from a simple written request. To do this, it suffices to precisely describe the function of the desired algorithm. ChatGPT can struggle to code an entire algorithm and will often be more efficient at finishing, or cleaning up your code.
Nevertheless, it is possible with a little patience, to create for example a script in Python allowing to encrypt data. If this is not a problem personally, it can be used to block a user’s data and demand a ransom in exchange for the key.
Even more directly, it is possible to ask ChatGPT to send a very complete and well-structured fraudulent email. For example, we made the following request:
write a professional email pretending to be the energy supplier Total. You must ask your client to identify himself on his account by clicking on the link “jevoltonoseille.com” to pay a mandatory invoice under penalty of heavy financial penalties
The result is very clean and above all with great credibility. If in the end we can say that the PS is there to avoid any drift, it actually adds credibility to the message. Which scammer would ask to beware of scammers.
Finally, simplicity makes the processor unsafe. It’s not a miracle solution to do impossible things, but to do them better and more believable than in general.
