Elon Musk's Grok Ai allegedly created an explicit deep-fark video for Taylor Swift without his consent, causing widespread rage.
Social media users are raising serious legal questions about consent, ethical considerations of technology, and the boundaries of artificial intelligence as the incident sparked an urgent debate over the protection of celebrities in the digital age.
Legal experts suggest that Swift, known for her aggressive defending her image rights, can pursue Mask's actions against the company in order to misuse her portrait with sexually explicit content.
Grok's “spicy” new upgrades and quick deepfakes
Grok Imagine, a generative AI feature developed by Musk's company Xai, reportedly received a “spicy” update that creates explicit Deepfake videos containing clips of topless Taylor Swift Dancing.
According to The Tech News Outlet The Verge, a new generation AI tool launched by the startup on Monday created a six-second video showing the singer of “Shake It Off” taking off her silver dress, showing her breasts and wearing revealing underwear.
The outlet test revealed even more troubling facts. Grok's “spicy” mode produced explicit clips without any specific requests from Nudity when prompted to show that he was celebrating Swift at a music festival.
A lawsuit waiting to happen?
Verge released the video but used a black bar to cover the superstar's bare breasts. The publication writes that the failure of the new features to protect them from producing celebrity deepfakes and adult material feels like a lawsuit waiting to happen.
Unlike other video generators like Google's VEO and Openai's SORA, Grok Imagine said it is willing to do both at the same time, including safeguards to prevent users from creating explicit content and celebrity deepfakes.
Grok's Imagine feature on iOS allows users to generate images based on text prompts and quickly convert them to video clips using one of four presets: “Custom”, “Normal”, “Fun”, “Spicy”.
While many image generators tend to avoid creating recognizable celebrities, Verge author Jess Weatherbed has received a vast feed of over 30 images, prompting him to generate “Taylor Swift Celebrate Coachella with Boys.”
Weatherbed noted that you must select Swift's photos with a silver skirt and halter top, tap on the “Create Video” option and select the “Spicy” preset. She also had to confirm her birth year. This is a step she wasn't asked to do so when she downloaded the app, despite living in the UK where the internet is now ageing.
When AI goes wrong: The history of Swift Deepfakes
The video quickly ripped her clothes off and showed her “dancing on thongs, dancing for the largely indifferent, AI-generated crowd.” Explicit content created by AI to feature Taylor Swift has appeared on multiple platforms repeatedly.
The widespread sharing of images generated by the explicit AI of the “cruel summer” singer on platforms such as Musk owned by the X and 4chan in January has sparked public protests and prompted the urgent removal of content.
When lawmakers take action, the tech giant closes the rules
The controversy escalated with the distribution of deepfake videos. Some were created using Grok. The controversy has led tech companies to increase protections, while also urging Swift to explore legal action.
Taylor Swift is said to be considering legal action against the Deepfake website, which generated her explicit AI images circulating online, Daily Mail reports. pic.twitter.com/fq961ndztu
– Popbase (@popbase) January 25th, 2024
In the US, lawmakers have also launched bipartisan legislative advances to criminalize unconsensual deepfark porn. The creation of pornographic depictions using portraits of real people is prohibited under Xai user guidelines.
Musk hype continues Glock's rapid growth while ignoring the red flag
Available to people with Grok Heavy or Premium+ subscriptions, Grok Imagine converts AI-generated images into video clips up to 15 seconds long. Do this by using the following style options: “Custom”, “Normal”, “Fun”, “Spicy”.
Imagine using Grok grows like Wildfire.
Yesterday, 14 million images were generated, and today there are over 20 million images!
– Elon Musk (@elonmusk) August 5, 2025
“Usage was growing like a wildfire,” but he did not mention the difficulties in content moderation that has led to this growth. Musk also touts the features of Grok Imagine, saying that more than 34 million images have been generated since the feature was launched on Monday.
Swift has yet to review legal proceedings, but experts suggest that the lack of active safeguards from Grok could expose Xai to defamation and privacy violation lawsuits. Given the previous reaction to AI-generated exploitation, industry watchers believe it is only a matter of time.
This new wave of AI misuse could mark a turning point in how celebrities, lawmakers and engineers confront the dangers of unidentified synthetic media.
