Musk’s AI chatbot faces global backlash over sexualized images of women and children

LONDON — Grok, Elon Musk’s AI chatbot, is facing a backlash from governments around the world after a recent wave of sexualized images of women and children generated without consent by the artificial intelligence-powered tool.

Britain’s top tech executive on Tuesday demanded that Musk’s social media platform X take urgent action, while a Polish lawmaker cited it as a reason to enact digital security laws.

The European Union’s executive body denounced Grok while officials and regulators in France, India, Malaysia and Brazil condemned the platform and called for investigations.

Growing concern from disparate countries highlights the nightmarish potential of nudification apps that use artificial intelligence to generate fake, sexually explicit images.

Here’s a closer look:

The problem emerged after last year’s launch of Grok Imagine, an AI image generator that lets users create videos and images by typing text prompts. It includes a “spicy mode” that can generate adult content.

This snowballed late last month when Grok, hosted on As of Tuesday, Grok users could still generate images of women using queries such as “put her in a see-through bikini.”

The problem is amplified both because Musk presents his chatbot as a bolder alternative to rivals with more guarantees, and because Grok’s images are publicly visible and therefore can be easily distributed.

The nonprofit group AI Forensics said in a report that it analyzed 20,000 images generated by Grok between Dec. 25 and Jan. 1 and found that 2 percent depicted a person who appeared to be 18 or younger, including 30 of young or very young women or girls in bikinis or see-through clothing.

Musk’s artificial intelligence company, xAI, responded to a request for comment with the automated response “Legacy Media Lies.”

However, X did not deny the existence of the embarrassing content generated via Grok. Yet it has consistently maintained in a post on its Security account that it takes action against illegal content, including child pornography, “by removing it, permanently suspending accounts, and working with local governments and law enforcement when necessary.”

The platform also repeated a comment from Musk, who said: “Anyone who uses Grok to create illegal content will face the same consequences as if they upload illegal content.” »

A growing number of countries are demanding that Musk do more to curb explicit or abusive content.

X must “urgently” resolve the problem, Technology Secretary Liz Kendall said on Tuesday, adding that she supported further review by Britain’s communications regulator Ofcom.

Kendall said the content was “absolutely appalling and unacceptable in decent society”.

“We cannot and will not allow the proliferation of these humiliating and degrading images, which disproportionately target women and girls. »

Ofcom said on Monday it had made “urgent contact” with X.

“We are aware of serious concerns raised about a Grok feature on X that produces images of undressed people and sexualized images of children,” the watchdog said.

The watchdog said it had contacted X and xAI to understand the steps taken to comply with UK regulations.

Under the UK’s Online Safety Act, social media platforms must prevent and remove child pornography content as soon as they become aware of it.

A Polish lawmaker used Grok on Tuesday as grounds for national digital security legislation that would strengthen protections for minors and make it easier for authorities to remove content.

In an online video, Wlodzimierz Czarzasty, speaker of parliament, said he wanted to become a target of Grok to highlight the problem, as well as call on the Polish president to support the legislation.

“Lately, Grok has been undressing people. He’s undressing women, men and children. We feel bad about it. Honestly, I almost wish this Grok would undress me too,” he said.

The bloc’s executive body is “well aware” that Grok is used for “explicit sexual content with child-generated images”, European Commission spokesman Thomas Regnier said.

“It’s not spicy. It’s illegal. It’s terrible. It’s disgusting. That’s how we see it, and it has no place in Europe. This is not the first time Grok has generated such production,” he told reporters on Monday.

After Grok distributed denialist content last year, according to Regnier, the Commission requested more information from Musk’s social media platform X. X’s response is currently being analyzed, he said.

Paris prosecutors said they were expanding their ongoing investigation into X to sexually explicit deepfakes after authorities received complaints from lawmakers.

Three government ministers alerted prosecutors to “blatantly illegal content” generated by Grok and posted on X, according to a government statement last week.

The government has also flagged issues with the country’s communications regulator over possible violations of the EU’s Digital Services Act.

“The Internet is neither a lawless zone nor a zone of impunity: sexual offenses committed online constitute criminal offenses in their own right and are fully subject to the law, in the same way as those committed offline,” the government said.

The Indian government on Friday issued an ultimatum to X, demanding that it remove all “illegal content” and take action against offending users. The country’s Ministry of Electronics and Information Technology also ordered the company to review Grok’s “technical and governance framework” and file a report on the measures taken.

The ministry accused Grok of “gross misuse” of AI and serious breaches of its safeguards and enforcement by allowing the generation and sharing of “obscene images or videos of women in a derogatory or vulgar manner in order to indecently denigrate them.”

The ministry warned that missing the 72-hour deadline would expose the company to greater legal problems, but the deadline passed without any public update from India.

Malaysia’s communications watchdog said Saturday it was investigating X users who violated laws banning the dissemination of “grossly offensive, obscene or indecent content.”

The Malaysian Communications and Multimedia Commission said it was also investigating online harm on X and would summon a representative from the company.

The watchdog said it had noted public complaints about the use of X’s AI tools to digitally manipulate “images of women and minors to produce indecent, grossly offensive or otherwise harmful content.”

MP Erika Hilton said she reported Grok and X to Brazil’s federal prosecutor’s office and the country’s data protection watchdog.

In a social media post, she accused the two men of generating and then publishing sexualized images of women and children without their consent.

She said X’s AI functions should be disabled until an investigation is conducted.

Hilton, one of Brazil’s first transgender lawmakers, decried how users could trick Grok into digitally altering any posted photo, including “swapping women and girls’ clothes for bikinis or making them suggestive and erotic.”

“The right to image is individual, it cannot be transferred through the ‘conditions of use’ of a social network, and the massive distribution of child pornography content (asterisk)gr(asterisk) by artificial intelligence integrated into a social network crosses all borders,” she declared.

__

AP writers Claudia Ciobanu in Warsaw, Lorne Cook in Brussels and John Leicester in Paris contributed to this report.