Earlier, the AI technique was used to fabricate pictures of Trump being taken into custody.
Making false photos to spread political misinformation in the lead up to the approaching presidential election in the United States has become easier than it has ever been before because to the proliferation of artificial intelligence systems that can swiftly make changed images and videos. According to Bloomberg, the answer that Midjourney provides for this problem could be to completely prohibit the use of political pictures. It has been alleged that David Holz, the CEO of Midjourney, informed customers during a chat session on Discord that the company is getting close to banning photographs such as those of Biden and Trump within the next year.
“I know it’s fun to make Trump pictures — I make Trump pictures,” he told users who were there at the session as they were in attendance. “Through an aesthetic standpoint, Trump is really interesting. In spite of this, it is probably best to just not do it at all; it is preferable to withdraw a little bit during this election. Let’s wait and see.” According to Bloomberg, in the past, individuals had utilized the artificial intelligence of the corporation to create deepfakes depicting Trump being detained. Following the widespread dissemination of those photos, as well as those infamous deepfakes depicting the Pope donning a coat inspired by Balenciaga, the business decided to terminate the free trials of its artificial intelligence image generator.
The organization already has regulations in place that prevent the formation of “misleading public figures” and “events portrayals” that have the “potential to mislead.” These rules are now in effect. Using the older version of Midjourney’s technology, Bloomberg was still able to generate modified photographs of Trump coated in spaghetti. On the other hand, the more recent version of the system refused to generate modified images of the former President. It is important to note that even if Midjourney decides to exclude photographs of prominent politicians, the company will only be doing so in order to prevent their platform from becoming the target of criticism and the focal point of attention during this election season. The use of artificial intelligence techniques in political misinformation campaigns or the dissemination of fake material with the intention of manipulating elections as a whole would not be prevented by this measure.
In addition, several technology businesses have taken measures to assist in the prevention of political disinformation, or at the very least, to assist in making it easier to recognize. While Meta is attempting to develop technology that can recognize and indicate whether a picture, video, or audio clip has been generated using artificial intelligence, ChatGPT will soon begin making use of DALL-E 3 to tag photos that have been created using the program.