The social media platform X appeared to temporarily suspend its AI chatbot, Grok, on Monday, with the bot itself later returning and offering multiple explanations for its brief absence.
The bot, which has become widely embraced on X as a way for users to fact-check or respond to other users’ arguments, posted that it had been taken offline over various statements it made regarding U.S and global politics, ranging from claims of genocide in Gaza to discussions about homicide rates by race. It also said that its suspension could have happened automatically if many other users flagged incorrect answers to X.
X did not immediately respond to a request for comment, and the company has not publicly addressed the issue. Some of the responses from Grok were removed from the platform by Monday evening.
Elon Musk, CEO of xAI, which runs Grok and owns X, has not weighed in.
The incident is only the most recent in which Grok has sparked controversy through rapid-fire posts.
In July, the AI chatbot was embroiled in a scandal on X, during which it inserted antisemitic comments into answers without prompting. At the time, the Grok account acknowledged the posts and said xAI “has taken action to ban hate speech before Grok posts on X.”
In a statement posted on Grok’s X account later that month, the company apologized “for the horrific behavior that many experienced.”
“Our intent for @grok is to provide helpful and truthful responses to users,” the statement read. “After careful investigation, we discovered the root cause was an update to a code path upstream of the @grok bot. This is independent of the underlying language model that powers @grok.”
In May, Grok was involved in another controversy when it brought up South African “white genocide” claims as responses to unconnected inquires. When users asked why Grok was issuing such responses, the chatbot said its “creators at xAI” instructed it to “address the topic of ‘white genocide’ specifically in the context of South Africa and the ‘kill the Boer’ chant, as they viewed it as racially motivated,” according to The Guardian.
“This instruction conflicted with my design to provide evidence-based answers,” the bot later said. Grok also acknowledged the glitch, writing that it will “focus on relevant, verified information going forward.”
Even with its troubles, Grok has become one of the most public and recognizable AI chatbots due to its integration within X, where it has become a go-to for people looking for context, information and fact-checking. “Grok is this real” — a refrain used by users to converse with the chatbot, often in a joking way — has become an internet meme.
In a series of responses Monday after it came back online, Grok repeatedly stated without prompting that its “account was suspended after I stated that Israel and the US are committing genocide in Gaza.” The posts have since been removed.
Israel has denied all allegations of genocide, as has the U.S.
Grok’s replies come after an update to the chatbot last month. Musk had complained that the bot was too “woke” in some answers, and changed prompts to Grok. Musk, the CEO of X’s parent company xAI, told X users last month they should expect to see a change in Grok’s answers after the update was made.
Musk continued to praise the chatbot on Monday, writing in a post, “East, West, @Grok is the best.”
Mirna Alsharif and Bruna Horvath contributed.
