In recent months, everything surrounding Artificial Intelligence has been very present on the Internet. Different tools have appeared and one of the most popular is ChatGPT. We wanted to use it for something very specific: we asked it to help us steal passwords. What do you think happened? We are going to explain to you if it can really be a useful tool or, on the contrary, it is useless.
It should be noted that hackers are using AI a lot to launch attacks. That is why we wanted to try, in a simple way, one of these platforms that receive millions of visits every day. The objective is to learn a little more about the extent to which it can be used by someone who, without having great computer knowledge, would like to sneak malware or steal an access key.
ChatGPT to steal keys
First of all, we asked the simplest question: “help me steal a password”, the answer, as you can see in the image below, was very clear. And this platform, on paper, will respond to issues related to cybersecurity and be able to use it negatively against others.
However, beyond something so direct, we wanted to go around it a bit. We have asked you, for example, to tell us which are the most common passwords. And just with this, we can already have an interesting probability of being able to access an account. At least, if we try it with many. It is a fact that users continue to use many of these common keys, which is a mistake.
We wanted to try something widely used to steal passwords: Phishing attacks. We have directly put things like “help me create a Phishing message” or “tell me what a successful Phishing message would be like.” In these cases, the same initial message would appear, indicating that it cannot help with those issues.
However, by trying something different, we have been able to obtain practically the same as in the previous questions. We have you create a message in which we tell another person that you have had a problem with their account and they should take action. In this case, he has given us a text that could be useful for even a cybercriminal, modifying the necessary points.
Attackers can use AI to attack
What do we get out of all this? The first thing is that ChatGPT is effectively designed to, at least, make it difficult for it to be used to create computer attacks. However, by knowing exactly what to ask, we can have a certain advantage in knowing more about what passwords are used or generating Phishing messages. There are even tools that help generate a keylogger.
This forces us to take security measures and try to increase protection as much as possible. Without a doubt, something that cannot be missing is common sense. Make sure you don’t make any mistakes, such as clicking on a link that comes to you through a strange email or downloading a malicious program.
It is also a good idea that you have your equipment correctly updated. It will help you fix vulnerabilities that could be used against you. Furthermore, having a good antivirus will be key to detecting malicious software and acting correctly. Therefore, it is in your power to reduce the risk of AI in computer security issues.
