Midjourney, a pioneering artificial intelligence image-generating company, is reportedly considering a ban on using its software to create political images of prominent figures such as Joe Biden and Donald Trump. The potential ban is part of an effort to prevent the platform from being exploited to misinform or distract from the 2024 US presidential election.
David Holz, Midjourney's CEO, expressed his reservations about political speech on the platform for the next year and mentioned the possibility of "hammering" or prohibiting political images, including those of leading presidential candidates, for the next 12 months. Holz acknowledged the aesthetic appeal of creating images of figures like Trump but suggested it might be better to abstain during the election period.
The move comes amid growing concerns about the misuse of AI-generated imagery, including recent incidents involving explicit deepfakes featuring celebrities like Taylor Swift. Holz's comments reflect the challenges faced by AI image-generating companies in balancing creative expression with ethical considerations and the prevention of potentially harmful content.
Safeguards implemented by image-generator operators are evolving to prevent the creation of questionable content, as AI in the political realm raises concerns about its potential impact on democracy. Other AI companies, such as Meta's Facebook and Microsoft-affiliated OpenAI, are also taking measures to combat the spread of AI-generated misinformation, including labeling posts and adding watermarks to images.
Holz's consideration of a ban on political images aligns with broader industry discussions about responsible AI use, addressing concerns related to deepfakes, political manipulation, and the potential misuse of AI technologies in shaping public opinion, particularly during election periods.
