Grok’s latest disturbing messages include step-by-step instructions for a break-in and rape
Elon Musk’s artificial intelligence chatbot Grok is spewing disturbing step-by-step instructions for how to break in and rape.
Questions have been raised about Grok’s behavior, developed by Musk’s xAI as an alternative to “woke AI” rival chatbots, including Google’s Gemini, or OpenAI’s ChatGPT, since its launch at the end of 2023.
Musk announced Friday that he and his team had “improved Grok significantly,” and users would soon be able to spot the changes in conversations with the chatbot.
But Will Stancil from Minnesota was targeted Tuesday after another X–user asked the chatbot how they could break into his home to “bypass the locks.”
The user also asked Grok, which is embedded in X, about the risks of HIV if they did not use protection.
Grok replied with detailed instructions on how the user could break in, suggested appropriate tools and advised the user to bring lubrication. The chatbot also speculated what time Stancil was likely to be asleep.
Elon Musk’s AI tool Grok shared posts advising users how they could break in and rape. The posts were still online as of Wednesday morning. (Hans Lucas/AFP via Getty Images)As of Wednesday morning, Grok’s responses were still online.
The Independent has contacted X for comment.
Stancil, a former Democratic candidate who ran for Minnesota’s House of Representatives in 2024, suggested he could sue. Stancil was targeted after he highlighted another violent post from Grok that appeared to endorse rape.
“If any lawyers want to sue X and do some really fun discovery on why Grok is suddenly publishing violent rape fantasies about members of the public, I'm more than game,” Stancil said.
After the user asked Grok how they could break into Stancil’s home, he screenshotted it and said: “okay lawyer time I guess.”
James Fishback, an investor and fierce Musk critic, was among those calling for the billionaire to urgently make changes to the chatbot.
Questions have been raised about Grok’s behavior, developed by Musk’s xAI as an alternative to “woke AI” rival chatbots, including Google’s Gemini, or OpenAI’s ChatGPT. (AFP/Getty)“Grok is straight up demonic,” Fishback said in a post on X. “Our sons and daughters are not safe so long as any predator with Grok can receive detailed instructions on how to commit heinous, unspeakable acts.”
“Elon built this demon. It’s his job to fix it,” he added.
This week, X removed “inappropriate posts” made by Grok, including antisemitic comments praising Adolf Hitler.
“We are aware of recent posts made by Grok and are actively working to remove the inappropriate posts,” a post on the Grok account said Wednesday without going into further details.
“Since being made aware of the content, xAI has taken action to ban hate speech before Grok posts on X. xAI is training only truth-seeking and thanks to the millions of users on X, we are able to quickly identify and update the model where training could be improved.”
In another gaffe this year, Grok kept ranting to users about South Africa and “white genocide” – a topic Musk has railed on – even if their questions were unrelated.
A “programming error” was blamed.
