The debate over the use of AI-generated content by US President Donald Trump and his administration was reignited this week when Trump posted a video depicting the arrest of former President Barack Obama.
The 1-minute-26-second video, postedon Trump's own Truth Social platform on Monday, opens with authentic clips of Obama and other senior Democrat politicians, past and present, declaring that "no one is above the law." It then cuts to footage appearing to show Obama being arrested by FBI agents in the Oval Office, handcuffed in the presence of Trump, and later shown behind bars in an orange jumpsuit.
Of course, Obama has not been arrested. The latter part of the video is clearly generated using artificial intelligence. It even bears a watermark revealing that it has been created by a TikTok account with only around 8,000 followers that regularly posts sensationalist, pro-Trump content, some of which is AI-generated.
Yet Trump still posted it to his 10.5 million followers, just as he did with a clearly fake image depicting him as the pope on May 3 and an AI-generated video of a "Gaza Riviera" on February 26.
The latter example sparked particular outrage, as it presents a vision for the Gaza Strip under the controversial proposal Trump put forward in February 2025. The US president suggested a full US "takeover" and redevelopment of the territory, aiming to transform Gaza into what he described as the "Riviera of the Middle East" and resettling the Palestinian population in the process. Numerous human rights organizations and international actors have sharply condemnedTrump’s plan, calling it a violation of international law and an attempt at ethnic cleansing.
Last year, AI-generated content also played a role in Trump's reelection campaign, from a video showing him dancing alongside tech billionaire Elon Musk, to images portraying Democratic candidate Kamala Harris as a Soviet-style "communist." He also shared demonstrably fake images of Taylor Swift fans wearing "Swifties for Trump" T-shirts, despite the pop star's public endorsement of Harris. DW Fact check has previously investigated the impact of disinformation on the US election, including cases involving AI-generated fakes.
Why does Trump keep sharing AI fakes?
The fact that the content is not only fake but often patently absurd doesn't seem to bother Trump or his supporters and allies, who continue to regularly post similar motifs. So, why do they do it?
"The reason Trump shares a lot of AI-generated content is the same reason that a lot of right-wing trolls share AI-generated content: to contribute to the 'meme-ification' of politics," explained Alex Mahadevan, director of MediaWise, a digital media literacy program at the Poynter Institute in Florida.
"The White House itself has made AI-generated images on social media a core element of its audience engagement strategy. It's for the shock factor, the engagement factor. They're just trying to stay relevant. And unfortunately, if you look at the engagement, it works really well."
Ahead of the 2024 presidential election, many observers feared that AI content would muddy the waters so that voters simply wouldn't know what to believe. Ultimately, however, according to Mahadevan, that's not quite what happened.
"What AI actually ended up doing was just creating a propaganda machine on steroids," he told DW. "You can tell right away that it's AI-generated. But it's not designed to deceive the viewer; it's designed to push a political message."
Trump's tactic: strengthen support, enrage the opposition
Such obviously fake content as the Obama video shared this week may not be designed to deceive or convince anybody, but it is intended to provoke: enraging the opposition while strengthening Trump's base.
"It's rage bait," said Mahadevan. "A way to 'own the libs' [liberals], to 'own' the other side of the political aisle. When they see the other side sharing a meme or content and getting upset about it, they like that! They're doing a touchdown dance when they see that."
It may seem infantile but, according to Mahadevan, it's an established form of communication.
"The new right, the digital right, the trolling right have become a pretty big portion of Trump's base and this is how they talk to each other," he says. "Cool, AI-generated images. They are speaking to their base."
How Trump benefits from the 'liar's dividend'
Nevertheless, whether Trump himself explicitly means it or not, his use of fake content does contribute to a socio-political phenomenon known as the "liar's dividend" — the idea that the prevalence of manipulated media fosters general skepticism, making it easier to dismiss real content as fake.
Trump capitalized on this during the election campaign, falsely claiming that a verified image of a large crowd greeting Kamala Harris off the vice-presidential plane ahead of a rally in Detroit had been altered using AI.
"[We are] entering the 'nothing is true and everything is possible' phase," commented Renée DiResta, a disinformation expert and former researcher at the Stanford Internet Observatory, on Threads at the time. "The ability to plausibly cast doubt on the real is the unintended consequence of being able to generate unreality."
Trump's vision of the world
There is, however, one final possible explanation for why Trump and his supporters continue to post content that is demonstrably fake: it reflects a vision of the world they ultimately hope could be real one day.
"These pieces of content amount to more than AI slop; they help to create a digital mirror world that reflects the future that Trump imagines, however preposterous it may seem," wrote The New Yorker in March. "In the real world, Trump's vision of Gaza as an ethnically cleansed luxury resort may seem like political fantasy. But, on the internet, Trump Gaza already exists as a virtual beachside destination to like and to share."
Mahadevan concurs, explaining: "Trump is affected by AI in the same way that regular people are. You can use AI to manifest a future or a version of yourself that you would dream of. So, I don't think it's too much of a leap to say that Trump and his acolytes are using AI to manifest how they feel the world should be."
Following his reposting of the fake "Swifties for Trump" T-shirts in 2024, Trump told Fox News: "I didn't generate them. These were all made up by other people. AI is always very dangerous in that way. It's a little bit dangerous out there."
But that hasn't stopped him from continuing to post such content.
