Midjourney CEO David Holz has reportedly announced restrictions on its AI image generator regarding election-related content. A move meant to prevent users from abusing the AI image generator, most especially during the upcoming elections.
Most notably, the ban comes with restricting users from creating images of Joe Biden and Donald Trump. During a digital office hours event on Wednesday, Midjourney CEO David Holz informed several hundred of the service's loyal users that, with the election well underway, it is time to step back from election-related matters for a while.
(Photo: JIM WATSON,BRENDAN SMIALOWSKI/AFP via Getty Images)
Texas leads legal battle against federal government over border policies, asserting state authority. Friction with Biden administration escalates, shaping high-stakes political showdown.
The restrictions come as Google recently imposed restrictions on its AI chatbot, banning Gemini from answering certain election-related queries and prompts.
The first to test the new policy was the Associated Press, which reported that users requesting an image of Trump and Biden would get a warning with the words "Banned Prompt Detected." Users were then notified that they had generated an abuse alert after making a second effort to generate an image similar to the one they had requested.
With Midjourney, users can create realistic AI-generated photos of almost anything, including purportedly fake pictures of Vice President Joe Biden grinning while brandishing a machine gun and former President Donald Trump meeting Pope Francis in the Oval Office.
Midjourney's Non-Political Sentiments
The AI Company, which reportedly employs just 11 people according to its website, has largely stayed silent in the public debate around the possibility that generative AI technologies could be used to disseminate misleading information about elections throughout the globe. In February, Midjourney, the sole producer of a well-known image-generating tool, rejected to join a voluntary tech industry association designed to combat deepfakes produced by AI that aim to mislead voters.
The restriction comes after the AI company hinted last month that it was thinking about prohibiting users from creating political graphics using its artificial intelligence software. Holz somewhat implied, however, that he does not care much about political speech within the Midjourney platform.
The CEO went so far as to claim that he understands the joy of creating images of Trump and that he enjoys doing so himself because he finds the former president to be aesthetically interesting.
AI Election Restrictions
The CEO of Inflection AI, Mustafa Suleyman, declared in October that no political candidate would be supported by the chatbot employed by his company. Last February, Meta Platforms Inc. announced that, as part of a larger initiative to avoid disinformation in an election year, it will start tagging more postings that were made with AI techniques.
Additionally, as part of previously stated steps to counter political misinformation, OpenAI began adding watermarks to photos created using its platforms last mid-February.
Bans across different AI platforms relating to the upcoming elections continue to increase, all to avoid election disinformation. Tech Times reported that the dissemination of false, harmful, and malicious content online that threatens to tarnish public perception and disturb the integrity of the electoral process is one of the biggest concerns of this election season.
