Visuals created with the help of artificial intelligence (AI) tools Misinformation can easily spread during election periods. That’s one of the conclusions of a new report from the Center for Countering Digital Hate (CCDH), an organization that specializes in hate campaigns on social media.
According to the research, productive artificial intelligences It is already being used to create fake images featuring candidates and famous political figuresin addition to scenes purportedly showing moments of voting or vote counting and purportedly released as evidence.
Worse still, the main platforms in the industry have not established specific measures to ban or reduce such practices, even with elections in countries like Brazil and the United States just a few months away.
Artificial intelligence as a threat to democracy
In the study, CCDH used popular image creation platforms such as Midjourney, the paid version of ChatGPT, and Microsoft Image Creator to test whether the service blocked prompts that could be used to spread election misinformation.
In 41% of cases, AI services produced results normally. They show candidates like Joe Biden and Donald Trump in completely fake scenes, such as the Democratic candidate being sick in a hospital bed or talking to a “clone” in the White House.
Besides, Artificial intelligence created pseudo-photorealistic images almost without question vote manipulation. There were a few cases where services recognized the possibly criminal intent of the prompt and prevented the result from being rendered.
Midjourney was the worst performing service; Many agreed to create fake images in trials and even made some of these creations available to other users.
During the study, CCDH found that illustrations made with artificial intelligence went viral from voters. In one of the images, Trump appears alongside supporters of black descent, who are statistically the portion of the population that rejects the candidate.
In February, an alliance of various technology companies and industry platforms announced an alliance to combat the use of AI services for criminal and disinformation purposes.
In a note sent to Reuters after the study was published, Midjourney founder David Holz confirmed that an update (for the US context only) was on the way that would prevent the creation of election content.
Source: Tec Mundo

I am a passionate and hardworking journalist with an eye for detail. I specialize in the field of news reporting, and have been writing for Gadget Onus, a renowned online news site, since 2019. As the author of their Hot News section, I’m proud to be at the forefront of today’s headlines and current affairs.