The rise of AI videos in politics: From deepfakes to viral, lowbrow humor

Since the arrival of the first images created with artificial intelligence (AI), there has been talk of their political danger. For now, perhaps due to the complexity of creating them and their obvious inaccuracies, they seemed to be readily accepted by the public. Even the announced risk of deepfakes, though very real, was met with some indifference. It remains to be seen what will happen when these videos become indistinguishable from reality.

But politics in general — and specifically the U.S. presidency — still has the power to give these videos a particular meaning: the ability to influence, entertain, and spark humor. U.S. President Donald Trump launched his unrealistic vision for the future of Gaza as a resort destination, which it has now been revealed was created by two Israeli-Americans just to test new software. And in Spain, the conservative Popular Party (PP) created a mock version of Temptation Island with characters from the governing Socialist Party (PSOE), which ended up offending the government of the Dominican Republic.

Beyond partisan politics, there are growing examples of how these types of videos are expanding their influence. The rise of AI-generated clips has given birth to a genre known as “point of view” or POV, designed to imagine what life would have been like in different historical periods. Hundreds of videos on Instagram and TikTok showcase scenarios such as what a builder of the Egyptian pyramids, a soldier in the U.S. Civil War, a citizen in the midst of the bubonic plague, or a worker at Chernobyl in 1986 would have seen. There are even clips of hell, which TikTok hides as “sensitive content.” Most of these videos present clear, sanitized, and gentrified versions of the past, as though being a child in Ancient Egypt was an idyllic experience. Their goal is to go viral, and images of hardship do not achieve that.

One of the most viral examples of this trend in Spain was the video titled “You wake up as a teenager in 80s Spain.” The footage depicts a world that no longer exists, yet didn’t exist even in that decade either. But nostalgia has immense sociopolitical power. This vision shows a supposedly prosperous Spain — free from drugs, cell phones, immigrants, and crime — that some political parties promote as an ideal. “Video games on Atari, arcade centers everywhere, family Christmas dinners, Nike sneakers, record and cassette stores, we copied everything we saw on TV, and it was great,” says one TikTok comment.

The video is adapted from a clip depicting what it was like to be a teenager in the 1980s in the U.S., which went even more viral. AI now allows us not only to imagine alternate realities, but to experience them with minimal cost or effort. In recent weeks, we’ve witnessed a moment similar to the one when the Pope was shown wearing a Balenciaga coat in March 2023, but this time in video form. While it may not be mistaken for reality, it showcases the immense potential of AI in video creation.

OpenAI’s video creator Sora is already available in Europe, but these clips are created in a more basic way: small scripts written with ChatGPT are then turned into images and converted into motion. YouTube is full of recent tutorials on how to create them, and TikTok accounts dedicated to these topics include links to paid courses on how to make them. The videos of Spain in 1980 and Trump in Gaza are more refined. The time it takes to perfect these videos is now measured in days or weeks, not months or years.

Their unapologetic use in politics was inevitable, says Xavier Peytibí, political communication consultant at Ideograma: “Just as citizens normalize the use of AI, parties do so as part of their political communication.”

He continues: “That’s why I don’t believe it’s a passing fad, but rather a growing trend. We know more and more, and most importantly, because AI-generated videos offer a cheaper and faster way of producing eye-catching visual content. They also have a surprise effect, which makes them incredibly viral and shareable.”

Trump’s video has already received a Palestinian response, but it did not achieve the same viral success: the message and the person delivering it still matter.

The fears surrounding AI’s ability to create believable, but fake, content have not yet materialized. But it could still happen. One of the creators of the Gaza and Trump video, Solo Avital, believes we could eventually experience something akin to Orson Welles’ 1938 War of the Worlds broadcast, which fooled listeners into believing a Martian invasion was happening. Now Avital believes something similar could occur with an AI clip of “Trump announcing that World War III has begun and that he has just launched a nuclear weapon against Russia.”

Creating high-quality content would no longer be difficult: “If we had to make this as a film, we’d spend a week on it, and you wouldn’t be able to tell if it was real or not.” Avital spent eight hours on the Gaza video. The best defense against this issue is public skepticism.

For now, conservative political parties are the ones most willing to utilize these resources. “Traditional channels of political propaganda now seem insufficient, as the weight of the conversation has shifted to social networks, where the most extreme circles and figures like Donald Trump occupy more and more space with hateful communication and disinformation, which can have a much greater impact,” says Carmela Ríos, a journalist and professor specializing in networks and disinformation.

Moderate parties, according to Ríos, are hesitant to connect in this way with certain segments, such as younger people, who consume more content on social networks: “It makes them react because it’s fun, rude, or outrageous.”

Organizations that believe entering the AI field is a good idea should consider two things first, says Ríos: “First, AI-generated content is not real, and this recourse to fiction (or lies) must be very carefully contextualized. Second, the overabundance. When AI-generated content becomes widespread, it can end up saturating or boring the public, so it will be more important for the campaign’s focus to be on creativity, empathy, originality, and listening.”

For now, it seems that AI-generated content could end up becoming an accepted and widespread form of entertainment, where politics is mixed with humor: “They allow for much faster, cheaper, and more adaptable production, which, in turn, enables immediate responses to current events,” says Peytibí. “Current affairs and quick-response content, if well-done and creative, always work very well in political communication. Especially when humor is used, as it can reach less ideological and less politically active audiences through sharing in personal circles such as WhatsApp.”

Trump has not only integrated memes and podcasts into politics. Technological references will continue to grow as more and more voters become familiar with concepts like ASMR (videos with pleasant, soft sounds or voices), and they may even laugh with Trump about how the sound of deporting chained immigrants at an airport could be considered pleasurable.

Sign up for our weekly newsletter to get more English-language news coverage from EL PAÍS USA Edition