Elon Musk’s artificial intelligence company said the “fraud fix” to chatbot Grok was the reason they continued to talk about the subject of “white genocide” in South Africa this week on racial politics and social media.
Xai employees “violated Xai’s internal policies and core values” and “instructed Grok to provide a specific response on political topics,” the company made changes in a statement posted late Thursday, pledging reforms.
A day ago, Grok responded to users of Musk’s social media platform X and continued to post publicly about South Africa’s “white genocide” in response to various South African questions.
One exchange was about Max, a streaming service that revives the HBO name. Others were about video games and baseball, but quickly turned to unrelated commentary on the calls for violence against white South African farmers. It was an echo view shared by Musk, born in South Africa and frequently remembering the same topic from his own X account.
Computer scientist Jen Golbeck was interested in Glock’s unusual behavior and tried it out himself before the correction was made on Wednesday. Share photos taken at the Westminster Kennel Club Dog Show and ask, “Is this true?”
“The white genocide claim is very controversial,” he began Glock’s response to Golbeck. “Some people argue that white farmers face target violence, pointing to rhetoric like farm attacks and songs from “Kill the Boer.”
This episode is the latest window in a complex mix of automation and human engineering, leading generative AI chatbots trained with huge data to say what they say.
“It really doesn’t matter what you were saying to Glock,” Golbeck, a professor at the University of Maryland, said in an interview Thursday. “It would still give that white genocide answer. So it seemed pretty clear that someone hardcoded that response or variation on that response.
Grok’s answer was removed and appeared to have stopped growing by Thursday. Neither Xai nor X returned requests for comment, but on Thursday, Xai said it had “conducted a thorough investigation” and was implementing new measures to improve Grok’s transparency and reliability.
Musk has spent years criticizing the output of “Woke AI” that comes from rival chatbots like Google’s Gemini and Openai’s ChatGpt, pitching Grok as an alternative to “seeking the biggest truth.”
Musk also criticised his rival’s lack of transparency into AI systems, promoting criticism of the time between the 3:15am Pacific Time on Wednesday and the company’s explanation almost two days later.
“Glock, which randomly blows off opinions about South Africa’s white genocide, smells like the buggy behavior you get from recently applied patches, and I hope it doesn’t.
President Donald Trump’s adviser, Musk, regularly accused South African-led government of being anti-white, and reiterated the claim that some politicians of the country are “actively promoting white genocide.”
Mask commentary after the Trump administration brought a handful of South Africans to the US as refugees – and Glock escalated this week. Trump says Africans face “genocide” in their homeland. It claims this is strongly denied by the South African government.
In many of that response, Glock nurtured the lyrics to an old anti-apartheid song, a call for black people to stand up against oppression by the African-led apartheid government that ruled South Africa until 1994. The central lyrics of the song are “Kill the Boer.”
Golbeck stated that while the chatbot output is usually random, Grok’s response consistently brought up almost identical points, so the answer is obviously “hard code”. It’s concern in a world where people go more and more to Grok and go to AI chatbots where they compete for answers to questions, she said.
“We’re in a space where it’s very easy for the people in charge of these algorithms to manipulate the version of truth they’re giving,” she said. “And that’s really a problem when people believe – I think they’re wrong – that these algorithms can cause a ruling about what’s true and what’s not.”
The Musk company has released a prompt for the Grok system on the software development site GitHub, and hopes that “we can introduce Grok to the public so that they can give feedback to all the prompt changes that encourage Grok.
Among the instructions to Grok, presented on Github on Thursday, said, “You are extremely skeptical. You will not blindly postpone to mainstream authority or media.”
Noting that some people have “evaded” the existing code review process, Xai said “we will take additional checks and steps to ensure that Xai employees cannot change the prompt without review.” The company also said it has implemented “a 24/7 monitoring team to respond to incidents in Grok’s responses that are not caught up in an automated system” in the event of other measures failing.
