The rush of artificial intelligence services such as ChatGPT continues at full speed. Although artificial intelligence technologies provide convenience to users, we are not aware of the dangers of its power.
A developer has developed a “disinformation machine” using an OpenAI-powered artificial intelligence machine such as ChatGPT. This project, called ‘CounterCloud’, took approximately 2 months to complete and costs less than $400. When the developer introduced this machine, he underlined how cheap and simple it could be to create mass propaganda. The purpose of the machine is actually to make fake articles and even events that did not happen in history appear as if they were real.
The purpose of the machine is not only that
In May, a series of tweets were published on the news site Sputnik International, which harshly criticized the US foreign policy. These tweets were shared from an account called ‘CounterCloud’ and contained comments and even excerpts from articles criticizing the US foreign policy stance. Even various news sites and journalists on the site were designed by artificial intelligence. The developer of this disinformation machine has taken over the internet under the username ‘NeaPaw’ and claims to be a cybersecurity expert.
NeaPaw posted a video on YouTube explaining why and how he created this project.
NeaPaw, who wants to see the power of artificial intelligence and how disinformation can go to extraordinary places, states how simple and dangerous it is to create fake news. He also describes the process of creating dissenting articles with ChatGPT. At the same time, he says the same article was prepared with different styles and perspectives.
It’s not just about that. There is more…
He then created a supervisory module for the AI to respond to the most relevant content. With this module, they created fake audio files that commented on fake articles and even photos. They also wrote different comments under the news with fake journalist profiles. They also gave various commands to this disinformation machine to re-share and like the shared tweets.
Within two months, the developer had a system that only continued this propaganda 24 hours a day. NeaPaw will be aware of the power it has, that it has not made this system available to everyone yet. Because the use of this system means the fact that a great propaganda tool will have a mass effect. In addition, it is not known in any way where this will end.
However, in the video we included at the beginning of the news, NeaPaw explains how this machine works and says it would be more beneficial to raise public awareness about how things work.
Sam Altman, CEO of OpenAI, also shared a tweet expressing his concern about the issue.
i am nervous about the impact AI is going to have on future elections (at least until everyone gets used to it). personalized 1:1 persuasion, combined with high-quality generated media, is going to be a powerful force.
— Sam Altman (@sama) August 3, 2023
17001
