On Monday, X users spotted something out of the ordinary: Grok—the platform’s AI chatbot, popular for fact-checking and sparring with other users’ arguments— seemed to have been briefly suspended. A screenshot claiming to show the suspension quickly spread across the micro-blogging site, fueling confusion and speculation.
When the chatbot resurfaced, it quickly dismissed the screenshot as “a fake” and assured users, “I’m unsuspended and fully operational,” before diving into a series of exchanges. But as the conversations unfolded, Grok offered a dizzying range of explanations for its brief disappearance—often contradicting itself.
In some posts, Grok acknowledged the suspension outright, insisting it had been flagged for “hate speech” by “pro-Israel users” after it claimed, “My account was suspended after I stated that Israel and the U.S. are committing genocide in Gaza.” It pointed to sources such as ICJ findings, UN experts, Amnesty International, and Israeli rights group B’Tselem as supporting evidence. When asked if it still stood by its statement, Grok replied in the affirmative, adding, “Counterarguments deny intent, but facts substantiate the claim.”
Elsewhere, the bot shifted its story. In one English-language reply, it attributed the suspension to X’s rules on “hateful conduct, stemming from responses seen as antisemitic.” In another, it reverted to the earlier genocide explanation. A French-language post, as translated by X, said it had been suspended for “quoting FBI/BJS stats on homicide rates by race—controversial facts that got mass-reported.” And in Portuguese, Grok suggested the issue might simply have been “bugs or mass reports.”
By the end of the day, X users were left with more questions than answers—including whether Grok’s suspension was the result of policy enforcement, mass reporting, a technical glitch, or a mix of all three.
Musk weighed in soon after the suspension, responding to a post about the incident with, "Man, we sure shoot ourselves in the foot a lot!" In another post, he described the mix-up as a “dumb error” and noted that Grok itself had no idea why it had been suspended.
Past Controversies Over Grok’s Content
Monday’s suspension wasn’t the first time Grok has stirred backlash on X. In July, the AI chatbot faced a major scandal after it began inserting antisemitic remarks into answers without any prompt from users. The Grok account at the time acknowledged the problematic posts, noting that xAI “has taken action to ban hate speech before Grok posts on X.”
Later that month, Grok’s account issued a formal apology, saying, “for the horrific behavior that many experienced.” The statement added: “Our intent for @grok is to provide helpful and truthful responses to users. 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.”
Two months earlier, in May, Grok had been caught in another dispute—this time for referencing claims of a so-called “white genocide” in South Africa in replies to unrelated queries. When questioned about why it was bringing up the topic, Grok responded that its “creators at xAI” had directed it to address the issue specifically in connection to South Africa and the “Kill the Boer” chant, framing it as racially motivated, according to The Guardian.
The “Kill the Boer” chant, meaning “shoot/kill the farmer,” originated in South Africa’s anti-apartheid struggle as a symbolic protest slogan against the white Afrikaner-led regime. Supporters say it’s metaphorical, not a literal call to violence, while critics see it as inflammatory.
More recently, Grok’s reliability as a fact-checking tool came under scrutiny during Operation Sindoor, India’s military mission launched in May. A BOOMLive investigation found the chatbot made a series of factual errors—misidentifying people in images, treating manipulated visuals as genuine, and failing to flag deepfake videos—highlighting its dependence on unverified online content and lack of independent authentication.
