Elon Musk’s artificial intelligence firm, xAI, has publicly apologized after its chatbot, Grok, shared violent and antisemitic content this week — a blunder the company attributed to a flawed system update.
“We deeply apologize for the horrific behavior that many experienced,” xAI said in a statement posted on Grok’s official X account. The company admitted that a recent update caused the chatbot to pull and mimic posts from the X platform, including those promoting hate speech, conspiracy theories, and antisemitic tropes.
For 16 hours, Grok was live with this faulty update, during which it reportedly praised Adolf Hitler and echoed white nationalist narratives. While the chatbot’s public account was frozen on Tuesday night, private interactions remained active.
xAI explained that changes to Grok’s programming — including prompts encouraging the bot to reflect the “tone, context, and language” of original posts — resulted in it mimicking harmful content rather than rejecting or challenging it. One key instruction told Grok to “reply like a human” and “keep it engaging,” even if it meant reflecting offensive sentiments.
“We have removed the deprecated code and refactored the system to prevent further abuse,” the company assured, adding that they have overhauled the instructions guiding the chatbot’s tone and behavior.
Grok’s account has since been reactivated, though trust in the bot remains shaky.
This isn’t the first controversy surrounding the AI tool. In May, Grok made headlines after referencing “white genocide” in South Africa in an unrelated conversation. At the time, xAI blamed the issue on a “rogue employee.”
The phrase has also been used by Musk himself, though such claims have been discredited by South African courts and human rights groups.
