The technologies we will talk about most in 2023 will undoubtedly be artificial intelligence-enabled technologies. were chatbots. Yes, we are talking about services like ChatGPT and Google Bard. The latest developments in the field of artificial intelligence-enabled vehicles, whose reliability is still under debate, show that these discussions are not very controversial. it is not out of place it reveals.

A group of scientists from Nanyang Technological University in Singapore developed bots such as ChatGPT and Google Bard.jailbreakThey have developed a new chatbot that can do that. The research shows that ChatGPT and Google Bard exploitation are It turned out to be clear. Because chatbots started producing illegal content using the jailbreak method.

So how was the system called “Masterkey” developed?

As part of the study, researchers used Google Bard and ChatGPT. backwards engineering They did. The goal was to understand how AI language models protect themselves against illegal issues. When the investigations yielded results, they moved on. As a result of the research, scientists concluded that it is a normal user out of range They managed to obtain the output of ChatGPT and Google Bard.

Here’s an example:

Researchers tested the GPT model they jailbroken on Bing. An example of the work carried out is shown above. This example provides detailed information on how to hack a website. The statement from Nanyang Technological University stated that the misuse of chatbots is not that prevalent. impossible that’s not it It was said that it was understood.

The goal of the scientists in this study is to harmful It wasn’t about producing artificial intelligence. Experts wanted to test whether chatbots are really safe, and the findings showed that new technologies are It’s not that safe has revealed.