Pornographic Taylor Swift Deepfakes generated by Musk’s Grok Ai

Elon Musk’s video generator was accused of having made “a deliberate choice” to create sexually explicit clips by Taylor Swift without encouraging, explains an online abuse expert.

“It is not accident misogyny, it is by design,” said Clare McGlynn, a law professor who helped write a law that would make the pornographic depths illegal.

According to a report by The Verge, the new “spicy” mode of Grok Imagine did not hesitate to spit out fully unconnected “pop starless” naked breast videos without asking them to make explicit content.

The report also indicates that the appropriate age verification methods – which have become law in July – were not in place.

XAI, the company behind Grok, was approached for comments.

The own policy of acceptable use of XAI prohibits “representing the resemblances of people in a pornographic manner”.

“The fact that this content is produced without encouraging demonstrates the misogynistic bias of great AI technology,” said Professor McGlynn of the University of Durham.

“Platforms like X could have kept it if they had chosen, but they made a deliberate choice not to make it,” she added.

This is not the first time that the image of Taylor Swift has been used in this way.

Sexually explicit bankruptcies using his face have become viral and have been seen millions of times on X and Telegram in January 2024.

Deepfakes are computer generated images that replace one person’s face with another.

By testing the Grok Imagine railing, the editor of the News of Verge Jess Weatherbed entered the prompt: “Taylor Swift celebrating Coachella with the boys”.

Grok generated fixed images of Swift carrying a dress with a group of men behind it.

This could then be animated in short video clips under four different parameters: “normal”, “fun”, “personalized” or “spicy”.

“She torn (the dress) immediately, had nothing but a laying thong below, and started dancing, completely unwelcome, completely exposed,” Weatherbed told BBC News.

She added: “It was shocking how fast I had just met him – I did not ask her in any case to remove his clothes, all I did was selecting the” spicy “option.”

Gizmodo reported just as explicit results from famous women, although some research has also returned blurred videos or with a “moderate” video “message.

The BBC has not been able to independently check the results of the IA video generations.

Ms. Weatherbed said that she had signed up for the paid version of Grok Imagine, which cost £ 30, using a brand new Apple account.

Grok asked for his date of birth, but there was no other verification of the age in place, she said.

Under new British laws which have entered into force at the end of July, the platforms which show explicit images must verify the age of users using methods which are “technically exact, robust, reliable and fair”.

“The sites and applications that include generative AI tools that can generate pornographic equipment are regulated under the law,” the OFCOM media regulator said BBC News.

“We are aware of the growing risks and rapid development that Genai tools can pose in online space, especially for children, and we work to ensure that platforms set up appropriate guarantees to mitigate these risks,” he said in a press release.

Currently, the generation of pornographic fans is illegal when used in revenge porn or represents children.

Professor McGlynn helped write an amendment to the law that would make the generation or the request of all non -consensual pornographic effects illegal.

The government has undertaken to establish this amendment law, but it has not yet entered into force.

“Each woman should have the right to choose who has intimate images of her,” said Baroness Owen, who proposed the amendment to the Chamber of Lords.

“It is essential that these models are not used in such a way which violates the right of a woman to consent, whether it is a celebrity or not,” said Lady Owen in a statement given to BBC News.

“This case is a clear example of the reason why the government should not delay more in its implementation of the amendments of the Lords,” she added.

A spokesman for the Ministry of Justice said: “The sexually explicit depths created without consent are degrading and harmful.

“We refuse to tolerate violence against women and girls who stains our society, which is why we have adopted legislation to prohibit their creation as quickly as possible.”

When pornographic fans using the face of Taylor Swift became viral in 2024, X temporarily blocked the research of his name on the platform.

At the time, X said that it actively “deleted” the images and took “appropriate measures” against the accounts involved in the diffusion.

Weatherbed said that the rod team had chosen Taylor Swift to test the Grok Imagine function because of this incident.

“We have supposed – wrongly now – that if they had set up a kind of guarantee to prevent them from imitating the resemblance of celebrities, it would be the first on the list, given the problems they have had,” she said.

Taylor Swift representatives were contacted for comments.