Why Is xAI Blocking Its Ex‑Engineer From Joining OpenAI?

Elon Musk's AI startup xAI has secured a court order temporarily prohibiting former engineer Xuechen Li from collaborating with OpenAI or any competitor on generative AI technology.

On September 4, 2025, Judge Rita F. Lin of the Northern District of California approved xAI's emergency request, issuing an injunction that requires Li to hand over all personal electronic devices and cloud accounts within three days for forensic analysis.

The order also bars Li from any AI‑related work at OpenAI until investigators confirm that all stolen data have been recovered and deleted. A hearing on extending the injunction is scheduled for October 7.

Li, one of xAI's original 20 employees, helped develop the Grok large‑model and sold nearly $7 million of xAI stock before allegedly copying confidential files to his personal devices after resigning on July 28, 2025, and accepting an offer from OpenAI to start on August 19.

According to court filings, Li attempted to conceal the theft by deleting browser history, system logs, and renaming/compressing files. xAI claims he also refused to provide passwords for key accounts that could reveal the full scope of the breach.

Li holds a Ph.D. in computer science from Stanford, a background that includes a Meta doctoral fellowship, internships at Google and Microsoft, and research under Geoffrey Hinton. He was also a core contributor to the Alpaca series of large models.

xAI’s legal action emphasizes protection of intellectual property rather than accusing OpenAI of corporate espionage, highlighting the high stakes of AI leadership in a competitive market.