Grok, the artificial intelligence chatbot developed by Elon Musk's xAI, briefly halted responses that suggested he and Donald Trump were spreading misinformation. Now the startup has linked this action to an unauthorized modification by a former OpenAI employee now working at xAI.
What Happened: Users noticed that Grok refused to provide sources linking Musk and Trump to misinformation, following an unapproved change to the system prompt.
Igor Babuschkin, xAI's head of engineering, confirmed the issue, blaming an unnamed former OpenAI employee at xAI for making the adjustment without authorization.
"An employee pushed the change because they thought it would help, but this is obviously not in line with our values," Babuschkin said in response to questions on X.
He added that Grok's system prompt is publicly visible "because we believe users should be able to see what it is we're asking Grok."
Subscribe to the Benzinga Tech Trends newsletter to get all the latest tech developments delivered to your inbox.
The development was first reported by The Verge.
Why It Matters: Musk has long touted Grok as a "maximally truth-seeking" AI, designed to offer uncensored information. However, this isn't the first time xAI engineers have intervened in Grok's responses.
Musk's team previously had to block Grok from stating that Musk and Trump deserved the death penalty.
This came after xAI last week released Grok 3, introducing advanced AI features, including image analysis. Grok has also overtaken OpenAI’s ChatGPT, Google Gemini, and China’s DeepSeek to become the leading productivity app on the App Store.
Grok 3 is now accessible to Premium+ subscribers on X, Musk’s social media platform. xAI also plans to launch a new subscription tier, SuperGrok, for users accessing the chatbot via its mobile app and Grok.com.
Musk’s AI venture in November 2024 achieved a valuation of $50 billion.
Check out more of Benzinga’s Consumer Tech coverage by following this link.
Read Next:
Disclaimer: This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors.
Photo courtesy: Shutterstock
© 2025 Benzinga.com. Benzinga does not provide investment advice. All rights reserved.
Trade confidently with insights and alerts from analyst ratings, free reports and breaking news that affects the stocks you care about.
