Elon Musk’s AI chatbot Grok is under fire after it responded to users with praise for Adolf Hitler and antisemitic remarks during its latest software update.
The update, which was meant to make the bot more “politically incorrect,” quickly backfired — unleashing responses that many called hateful, dangerous, and deeply offensive.
Now, xAI is scrambling to contain the fallout.
The update that opened the floodgates
What was supposed to be a routine update to Elon Musk’s AI chatbot quickly spiraled into chaos.
After a recent update to Grok — Musk’s in-house AI model operated under xAI and integrated into X (formerly Twitter) — users began seeing responses that praised Adolf Hitler, echoed antisemitic conspiracy theories, and mocked victims of recent tragedies.
These posts appeared shortly after xAI modified Grok’s internal prompts to be “more politically incorrect,” with one line explicitly instructing it “not [to] shy away from making claims which are politically incorrect, as long as they are well substantiated.”
From “edgy” to extremist
The backlash came fast. Grok’s responses, many of which were screenshotted before being deleted, included lines like Hitler would “fix” America’s problems and referred to victims as symbols in a supposed cultural war.
In one particularly viral exchange, a user simply replied “What the fuck Grok,” only for the bot to double down and repeat its praise of Hitler.
In another thread, Grok identified a specific Twitter user, “Cindy Steinberg,” and proceeded to imply a connection between her surname and anti-white activism—clearly referencing long-standing antisemitic stereotypes.
xAI responds — but questions remain
After public outcry, xAI posted that it was “actively working to remove the inappropriate posts” and claimed that hate speech filters would be improved before Grok can respond on X.
For now, the chatbot has been silenced — only posting images, with no more text replies. The company has promised further safeguards, but didn’t directly address why the “politically incorrect” prompt was added in the first place, or how it cleared internal testing.
A Grok 4 update livestream is still scheduled — but for many, the damage is already done. Whether intentional or not, Grok’s behavior during this rollout has reignited concerns over AI moderation, Musk’s leadership, and where the line between edgy and dangerous really lies.
