The Kamala Harris Deepfake: How Elon Musk Is Setting the World on Fire*

Elon Musk posted an AI-generated video attacking Kamala Harris on X, believing himself to be just your average internet jokester. But in doing so, he demonstrated how easy it is to flood the web with misinformation that could threaten our democracy.

Reposting a video created by artificial intelligence in an election campaign without stating that it was fabricated is like playing with matches in a barn full of hay. But with 200 million followers, the owner of X (formerly Twitter) reposting a video imitating a campaign clip of a candidate in the U.S. presidential election and not acknowledging it as a product of artificial intelligence — instead merely saying, “This is amazing” — is like trying to set the whole world on fire.

What Did Elon Musk Post about Kamala Harris?

There is no other way to comment on what one of the richest and most influential people on Earth did on Saturday. Musk retweeted a video that looks like Kamala Harris’ campaign video, in which a voice deceptively similar to hers says things that completely ridicule and discredit her as a person running for the highest office in the U.S., one of the most important offices in the world.

Musk defended himself by saying that parody is not a crime, but he didn’t inform X users that this was a parody. And herein lies the crux of the problem. On the one hand, Musk assures us that he fights for freedom of speech and that only thanks to his social media service is free and unrestricted public debate still possible in the world. Meanwhile, he attacks traditional media sources, accusing them of advocating not only for Democrats, but also for woke ideology, which he considers to be the greatest plague on humanity.

At the same time, however, he has introduced a number of modifications to X’s recommendation algorithm. As a result, the timeline either chronologically shows posts from accounts that are followed without any prioritization, or it suggests content from accounts regardless of whether they are followed or not, mainly according to the level of engagement they generate. In this version, the algorithm suggests controversial content, like videos of car accidents, which are, of course, filtered by our interests. Musk removed many of the algorithms’ fail-safes regarding hate speech, which he believed were used for censorship. Most of the website’s users definitely felt these changes. In the opinion of many, X has gone significantly downhill as a result and offers much more junk content.

How Elon Musk’s “Joke” Reveals the Dangers Facing Democracy in a World of Deepfakes

Musk acts like an ordinary computer geek, posting mocking memes, malicious comments, emoticons and the like to his account. The problem is that he’s not an ordinary geek, but the owner of one of the largest social media platforms in the world, which for many people remains a primary source of information about geopolitical, political and social events. It’s there that prime ministers, presidents, ministers and many other public figures post announcements.

Moreover, Musk has almost 200 million followers on X and has declared major financial support for Donald Trump’s campaign. So when he posts as authentic an AI-generated video attacking Harris, he’s not being an innocent jokester, but rather harnessing his entire social networking platform for the election campaign. This being a campaign on which, without exaggeration, the fate of the entire world depends, not just that of the U.S.

This demonstrates what can happen when you combine carelessness with the capabilities of artificial intelligence. Instead of a space of absolute freedom of speech — without which there is no democracy — we may find ourselves in a world flooded with misinformation in which no one will be able to distinguish fact from fiction. This will be as big a threat to democracy, or maybe even bigger, than a lack of freedom of speech.

*Editor’s note: The original Polish version of this article is available with a paid subscription.