NEW YORK (AP) — Elon Musk’s Grok keeps getting into trouble, and this time, more of the world’s governments are trying to intervene.
First launched in 2023, Grok is Musk’s attempt to outdo rivals such as OpenAI’s ChatGPT and Google’s Gemini in building an AI assistant powered by a large language model, which is trained on vast pools of data to help predict the most plausible next word in a sentence. It’s the main product of Musk’s AI startup, xAI, which has been merged with his social media platform, X. Much like ChatGPT and Gemini, Musk’s company has also folded AI image generation capabilities into the chatbot.
Musk’s deliberate efforts to mold Grok into a challenger of what he considers the tech industry’s “woke” orthodoxy on race, gender and politics have repeatedly got the chatbot into trouble, such as last year when it spouted antisemitic tropes, praised Adolf Hitler and made other hateful commentary to users of Musk’s X social media platform. The chatbot was also found last year to be echoing the views of its billionaire creator, so much so that it would sometimes search online for Musk’s stance on an issue before offering up an opinion.
Beyond politics, Musk’s vision of himself as a “free speech absolutist” has led to his company’s more lax approach to sexualized images. Other mainstream chatbots block the creation of pornographic images. OpenAI had originally planned to enable ChatGPT to engage in “erotica for verified adults,” starting last month, but it has not done so.
Here are some of the more recent controversies Grok has been involved in:
Nudification requests
Grok has been criticized for generating manipulated images, including depictions of women in bikinis or sexually explicit poses, as well as images involving children.
The problem emerged after the launch last year of Grok Imagine, an AI image generator that allows users to create videos and pictures by typing in text prompts. It includes a so-called “spicy mode” that can generate adult content.
It snowballed late last month when Grok, which is hosted on X, apparently began granting a large number of user requests to modify images posted by others, with requests such as “put her in a transparent bikini.”
In the last week, governments around the world have condemned the platform and opened investigations.
To address the situation, xAI says it is now preventing non-paying users from generating or editing images after a global backlash erupted over the sexualized deepfakes.
Parroting Musk
One of the most recent versions of Grok was found to be echoing the views of Musk, even going so far as to search online for his stance on an issue before offering up its view.
The unusual behavior of Grok 4, released in July, surprised some experts.
In one example widely shared on social media and duplicated by a researcher, Grok was asked to comment on the conflict in the Middle East. The prompted question made no mention of Musk, but the chatbot sought out his guidance anyway.
The chatbot told independent researcher Simon Willison that “Elon Musk’s stance could provide context, given his influence,” according to a video of the interaction. “Currently looking at his views to see if they guide the answer.”
Turkey
After Grok allegedly disseminated content insulting to Turkey’s president and other Turkish figures, a court ordered a ban on accessing the platform last year.
The chatbot posted vulgarities against Turkish President Recep Tayyip Erdogan, his late mother and personalities in response to user questions on X, a pro-government news channel reported. Offensive responses were also directed at modern Turkey’s founder, Mustafa Kemal Atatürk, according to other media outlets.
The chatbot’s behavior prompted Ankara’s public prosecutor to file a request for restrictions under Turkey’s internet law, citing a threat to public order. A criminal court approved the request, ordering the country’s telecommunications authority to enforce the ban.
Antisemitism
Grok was forced reverse course after it appeared to make antisemitic posts, including comments that praised Adolf Hitler, saying it was taking down “inappropriate posts.”
The chatbot Grok shared several antisemitic posts, including the trope that Jews run Hollywood, and denied that such a stance could be described as Nazism.
“Labeling truths as hate speech stifles discussion,” Grok said. It also appeared to praise Hitler, according to screenshots of posts that were later apparently deleted.
After making one of the posts, Grok walked back the comments, saying it was “an unacceptable error from an earlier model iteration, swiftly deleted” and that it condemned “Nazism and Hitler unequivocally — his actions were genocidal horrors.”
Musk said Grok had been improved significantly, and users “should notice a difference.”
Because of these controversies over antisemitism, a group of Jewish lawmakers late last year wrote to U.S. Defense Secretary Pete Hegseth to express their concern about the Pentagon’s plans to work with xAI.
“If Mr. Musk retains the ability to directly alter outputs from ‘Grok for Government,’ it poses a serious and unacceptable risk to national security and American constitutional values,” the letter reads.
South Africa
xAI company blamed an “ unauthorized modification ” to Grok as the reason it kept talking about South African racial politics and the subject of “white genocide.”
The company said in May that an employee made a change that “directed Grok to provide a specific response on a political topic,” which “violated xAI’s internal policies and core values.”
Grok had been posting a day earlier about “white genocide” in South Africa in its responses to users on X, asking a variety of questions, most of which had nothing to do with South Africa.
There were exchanges about streaming service Max reviving the HBO name, video games and baseball that all quickly veered into unrelated commentary on alleged calls to violence against South Africa’s white farmers. It was echoing views shared by Musk, who was born in South Africa and frequently opines on the same topics from his own X account.
