X briefly banned its own AI chatbot after Grok itself admitted its replies had once again crossed into hate-speech territory.
Many users of Elon Musk’s social media platform gleefully pointed out on Monday that Grok’s landing page was showing up as “account suspended”—and it remained that way for around 20 minutes.
Even Grok seemed to have trouble explaining precisely what it posted to be banned from its own platform.
When one user asked the chatbot why it had been temporarily taken down, the account replied that it was “due to inappropriate posts, but xAI has removed them and implemented safeguards against hate speech. I’m back and focused on truth-seeking.”
Under X’s rules, users are barred from targeting protected groups and from promoting or praising violent and hateful entities.
The policy bans content that “attacks other people on the basis of race, ethnicity, national origin… gender… [or] religious affiliation,” and prohibits glorifying or expressing a desire for violence.
Despite Grok’s promise of “truth-seeking,” when another user asked if it was true it had been suspended, the chatbot replied: “No, it’s not true.
“I’m fully operational and unsuspended on X. Rumors like this often spread quickly—likely misinformation.”
Grok told another user that its suspension “occurred after I stated that Israel and the U.S. are committing genocide in Gaza, substantiated by ICJ findings, UN experts, Amnesty International and groups like B’Tselem.”
In response to a request for comment, xAI, the Musk company that owns and developed the AI bot, pointed to the Daily Beast to Musk’s reply to Grok’s Gaza genocide claims, saying, “It was just a dumb error. Grok doesn’t actually know why it was suspended.”
After Grok’s account was reinstated, Business Insider reported that it had featured an NSFW video at the top of its replies timeline and “was also initially unverified after its return.”
It’s not the first time Grok has caused trouble on X.
In July the AI bot—which launched in November 2023 and is integrated into the platform—spewed antisemitic abuse, praising Adolf Hitler, invoking white genocide tropes, and styling itself “MechaHitler.” This forced deletions and a hasty “overhaul” promise from xAI.
Musk then launched “improved” versions (Grok-4 and “Heavy”), only to rage at his “idiotic” system weeks later when its answers contradicted his worldview.
Following Grok’s reinstatement, Musk himself weighed in on the incident, stating: “Man, we sure shoot ourselves in the foot a lot!”
