Alex Brandon/AP
In much of the world, Grok and its parent company both appear to be in serious trouble. After Grok, X’s AI chatbot, has been used to generate sexualized and violent images of women and children, the social media company has faced a wave of backlash and censure, with new nationwide bans on accessing Grok in place and other consequences on the way. On Monday, the EU threatened to fine X under its broad Digital Services Act if it didn’t act “quickly” to fix Grok, in the words of one regulator.
But within the secure confines of Elon Musk’s own mind, Grok is not only wildly successful, but “solid as a rock,” as he tweeted on Monday, with the goal of pursuing “the deeper truth and appreciation of beauty.”
Musk slammed Grok’s critics as people who “just want to suppress free speech.”
The outrage over Grok’s lack of guardrails has been raging for weeks, after X delivered an update on Christmas Eve which allowed users to edit images and videos on the app. That included other people’s images and videos, and the tool was inevitably and quickly used for violently sexualized campaigns against the platform’s female users. Images of women in street clothes were undressed and images of women in hijabs and other religious modesty garments had them removed, as users generated images of women and girls bloodied and bruised. X eventually limited the tool to paid subscribers, which seems to mean that you can still make nonconsensual images if you pay for the privilege.
The Internet Watch Foundation, a UK non-profit, says it has uncovered “criminal imagery” of minor children generated by the tool. The UK’s Office of Communications, or Ofcom, the country’s regulatory body for communications services, says it has opened a formal investigation into X over the generation of nonconsensual images. In a statement reprinted by the Guardian and other outlets, Ofcom said it found reports “of Grok being used to create and share illegal nonconsensual intimate images and child sexual abuse material… deeply concerning,” adding that platforms “must protect people in the UK from content that’s illegal in the UK.” Citing the “risk of harm to children,” Ofcom pledged an investigation “as a matter of the highest priority.” Other countries have already acted: over the weekend, the governments of Malaysia and Indonesia both suspended access to Grok because of the generation of nonconsensual sexual images.
As he has in the past, Musk has largely stayed out of the discussion involving Grok, his most hideous and beloved child. But he has occasionally mounted public defenses of the AI platform; on January 3, he told a user, “Anyone using Grok to make illegal content will suffer the same consequences as if they upload illegal content.”
Overall, Musk has suggested that criticism of Grok is from people who “just want to suppress free speech,” and his broader strategy seems to be trying to focus on what he seems to see as the tool’s artistic value. Since the newest version of the tool was made available, Musk has shared several images and videos made by Grok that he considers to be beautiful or impressive, providing an interesting window into his definitions of those terms. They include a heavily made-up woman singing for four seconds, modelesque women dancing in dresses made of flowers, and a near-mute combination of those two things. He also reshared faked videos of a dragon swimming and a guy doing parkour on top of a tall building. Most recently, he reposted a Grok video of a busty, armor-clad vaguely Grecian-looking figure striding through stone ruins, declaring it to be “beautiful.”
Musk has also imputed sentience and motivations to Grok, taking on a loopy, quasi-religious tone when he discusses the tool. “Compared to other AIs, Grok is solid as a rock,” he declared in a tweet on Monday. “And it will get much better. Eternally curious to know the deeper truth and appreciation of beauty are its goals.” He’s previously said that Grok is “on the side of the angels.” As part of his continued and oft-stated ambitions to have X be the one, global company that everyone uses for everything, he’s also claimed that Grok and Optimus, the “humanoid robot” Tesla has claimed it is developing, could take the place of “government healthcare,” as Musk tweeted on Christmas. “Government healthcare is like having the DMV as your doctor,” he wrote. “Grok and Optimus will provide incredible healthcare for all.”
Meanwhile, in the more earthbound realities in which the rest of us live, one of the harshest critics of Grok to emerge on X is the mother of Elon Musk’s 13th known child, conservative influencer Ashley St. Clair, who has tweeted that she’s had “countless” explicit images made of her by the tool’s users. Her account’s blue verification checkmark was removed when she complained, she has claimed, and she says she was then banned from subscribing to Twitter Premium, the service that allows accounts to be monetized and subscribers access to what the company calls “increased usage limits” on Grok. (In a separate but related bit of messiness, Musk claimed he would file for sole custody of the pair’s son, accusing St. Clair of “planning to transition” the child after St. Clair apologized for past insensitive and inflammatory statements she’s made about transgender people.)
“I’ve had women send me the most horrid images Grok has produced of them,” St. Clair tweeted on January 6. “And the emails from the platform saying there was no violation of ToS [terms of service]. These women are distraught + at a complete loss for what to do.”
