Skip to content
IFAI

Industry

Trump shares fake AI-generated Taylor Swift endorsement

Published Reach 42

Donald Trump just shared a series of AI-generated images on social media, including a fake endorsement from pop star Taylor Swift alongside the caption "I accept". Legal experts note that current election deepfake laws struggle to police such exaggerated posts, though Swift herself could potentially pursue a right-of-publicity claim.

Trump shares fake AI-generated Taylor Swift endorsement
Image: theverge.com

If you saw Taylor Swift endorsing Donald Trump on your timeline this weekend, you were looking at generative AI. And the person sharing it was Trump himself.

Over the weekend, the former president posted a carousel of apparent AI-generated images to rally his base. The most brazen of the bunch showed pop superstar Taylor Swift dressed as Uncle Sam, pointing forward alongside text that read, "Taylor wants you to vote for Donald Trump." Trump’s response to the compilation? "I accept!"

He didn't stop there. Another post featured an image resembling Vice President Kamala Harris addressing a Chicago crowd ahead of the Democratic National Convention, with a massive communist hammer and sickle looming over the stage. Just days earlier, Trump had falsely claimed Harris used AI to fabricate crowds at her rallies. Now, he's the one leaning entirely into generated media.

Why election laws might not touch this

While around 20 U.S. states have passed laws targeting election deepfakes, Trump’s posts might easily slip through the cracks, according to Robert Weissman, co-president of Public Citizen.

Most state statutes only prohibit depictions that are convincing enough to be plausible. An over-the-top image of Harris under communist iconography is clearly caricature, making it hard to prosecute as an illicit deepfake.

At the federal level, guardrails are essentially nonexistent. Beyond an FCC rule against AI voices in robocalls, candidates face virtually no national restrictions on using synthetic media to distort reality. Under long-standing First Amendment precedents, political candidates have broad leeway to lie, and courts rarely step in unless an opponent can prove direct, quantifiable injury to voters.

Taylor Swift herself might have a stronger legal case than any election watchdog. Under state right-of-publicity statutes, such as California’s, celebrities have strong protections against unauthorized commercial or endorsement use of their likeness. Representatives for Swift and Universal Music Group haven't commented publicly on the matter, and the Trump campaign has remained silent.

Platforms look the other way

Private tech platforms have rules against deceptive synthetic media, but enforcement is erratic at best.

X has an explicit policy prohibiting manipulated media likely to cause confusion or harm. Yet enforcement has been notoriously selective, especially after platform owner Elon Musk reposted an unlabeled deepfake of Harris weeks ago. Meanwhile, on Trump's own Truth Social network, restrictions around synthetic media are practically non-existent.

Weissman notes the broader danger is Trump using AI to destroy the concept of objective reality entirely: claim real things are fake, share synthetic junk, and blur the lines until voters trust nothing they see with their own eyes.

Why it matters

We think this moment signals a permanent shift in how political campaigns weaponize AI, moving from covert deception to blatant, bad-faith trolling.

For everyday voters, the cognitive burden of using the internet just spiked again. The danger isn't that millions of people genuinely believe Swift became a MAGA supporter overnight. The real harm lies in normalizing fake media in mainstream political discourse. By embracing synthetic images, politicians can manufacture excitement, signal approval from figures who oppose them, and later hide behind the defense that "it was just satire."

Legally, it exposes a massive loophole in current legislation. Regulators have focused heavily on hyper-realistic deepfakes that trick the eye, but Trump’s posts show that cartoonish, half-believable AI content can be just as disruptive. While federal agencies move at a snail's pace, bad actors are exploiting this gap in real time.

Ultimately, platforms won't save us. With leadership at X and Truth Social showing little appetite to police their prominent users, the burden of truth falls squarely on the public. If people can no longer agree on what is real and what is generated, functioning democratic debate breaks down completely.

Sources (1)

  1. [1]
    theverge.com(opens in a new tab)Originaltheverge.com

Best of AI, daily