xAI Developer Fired for Spilling the Secrets of Grok 3

Benjamin de Cracker claims he resigned after xAI management pressured him to remove his post. He insists that his ranking of AI models was a personal assessment that neither disclosed confidential details nor breached company policies. Before leaving, he had spent six months in the Human Data department, which specializes in developing the Grok models.

His ranking positioned three OpenAI models—o1-pro, o1, and o3-mini—at the top, followed by Grok 3 in second place. Google’s Gemini 2.0 Pro Series landed at the bottom. xAI considered the post a violation of internal confidentiality and warned that keeping it online would result in termination.

Why did they tell me to remove this opinion? Well, according to them, the reason is that I acknowledged that Grok 3… exists. I wish I was joking. I'm not. The entire situation has been very strange. I thought about just deleting the damn thing…. But you know, once you start caving and giving up holding mild personal opinions, the slope becomes very slippery,

Benjamin De Kraker expressed.

He expressed disappointment that a company advocating for openness and free speech would take such drastic action against an employee for voicing a personal opinion.

Nevertheless, he reiterated his support for xAI’s broader mission, acknowledging its ambition to push AI development forward.
Elon Musk had previously stated that Grok 3’s training was already complete.

Surprisingly, not everyone saw Benjamin as the wronged party—some users sided with xAI, claiming that his post harmed the company’s reputation and could impact the release of Grok 3 later this year. His departure also takes place against the backdrop of a deepening conflict between Elon Musk and OpenAI CEO Sam Altman.

Read on: Elon Musk’s Grok Chatbot Now Understands Images