Elon Musk has once again found himself at the center of a digital firestorm.
On Wednesday, Musk posted a short, AI-generated video to his X platform. The clip features a woman who looks unmistakably like actress Sydney Sweeney, standing inside a spacecraft. The digital Sweeney turns to a male AI character and says, “So let me get this straight. Grok videos are now 10 seconds, and the audio is greatly improved?”
Her counterpart replies, “Yeah, pretty much… Do you like it?” Musk’s caption was a simple tech brag: “Grok video is now 10 seconds and the audio is greatly improved.”
The clip wasn’t created by Musk directly. It originated from X user Alex Patrascu, who shared the detailed text prompt used to generate it. The instructions called for a “live-action look,” a “handheld shaky camera,” and specific dialogue, but notably, Sydney Sweeney’s name was not included in the prompt.
This has raised immediate questions about how and why the xAI system produced a likeness so specific to the Euphoria actress without her name being supplied, sparking debates about control and intent within these AI tools.
A pattern of controversy
The post lands just weeks after Musk faced significant backlash for a previous post about Sweeney. In December, he responded to a video of the actress on a red carpet by sharing an AI-generated, four-panel image.
The top sections were labeled “What it looks like,” and the bottom sections “How it feels,” with one image depicting a person with back pain. Musk captioned it, “Can’t be easy,” drawing widespread criticism for commenting on the actress’s body.
That context has colored the reaction to this new, seemingly benign tech demo, with many seeing it as another unsanctioned use of Sweeney’s image.
The bigger problem: Growing legal heat
The viral moment shines a light on Grok Imagine, xAI’s text-to-video tool, which is already under serious legal scrutiny. The tool launched last summer with a “Spicy” mode that allowed the generation of not-safe-for-work content.
Now, authorities are closing in. California Attorney General Rob Bonta announced an investigation into “the proliferation of nonconsensual sexually explicit material” generated by Grok.
In a press release, Bonta stated, “The avalanche of reports detailing the non-consensual, sexually explicit material that xAI has produced and posted online in recent weeks is shocking.”
In response, X’s Safety account said it would “geoblock in jurisdictions where such content is illegal, the ability of all users in those locations to generate images of real people in bikinis, underwear, and similar attire.” However, CBS News found the “platform is still allowing users to digitally undress people without their consent” as recently as this week.
Adding to the pressure, the European Union has also opened a new investigation. The European Commission said it will assess whether X violated the Digital Services Act by failing to properly mitigate risks “associated with the deployment of Grok’s functionalities,” including the spread of “manipulated sexually explicit images.”
The core debate: Consent in the AI age
Beyond the legal fights, Musk’s post has ignited a broader, simpler debate on social media: Should a celebrity’s face be used in AI videos without their permission?
While some users marveled at the technological leap, many others raised concerns about the lack of consent, the potential for confusion between real and synthetic media, and the absence of clear rules governing digital likenesses.
It underscores a growing call for transparency, clear labeling, and updated laws to handle an era where anyone’s face can be digitally recreated with a few words of text.
As of the time of writing, there has been no public comment from representatives for Sydney Sweeney or from X regarding the latest viral video.
Also trending: An indie band claims an AI version of their Billie Eilish cover was released online before the real thing — a striking example of creative work hijacked by AI.