On August 30, 2025, Elon Musk and the artificial intelligence company xAI he founded filed a lawsuit against a former engineer in the California federal court. The core allegation of the lawsuit is trade secret theft, and the alleged target is Xuechen Li, a former employee of xAI.This person’s experience can be called a walking history of AI in Silicon Valley: he received a scholarship from Meta during his Ph.D., worked for Google and Microsoft, stole business secrets from xAI, and finally jumped to OpenAI.What is incredible is that these experiences combined did not exceed 4 years.


01

According to public information, Xuechen Li graduated from Beijing No. 4 Middle School in 2014. Subsequently, he studied at the University of Toronto in Canada from 2015 to 2019, receiving a Bachelor of Science degree in computer science, mathematics and statistics, and graduated with honors and received the Chancellor's Graduation Scholarship.

In 2021, he entered Stanford University in the United States to pursue a PhD in computer science. He completed his doctoral thesis defense in three years, and during his PhD study, he received a scholarship specially provided by Meta for doctoral students in the field of AI.

Xuechen Li received her PhD in 2024. His doctoral research advisors are Tatsunori Hashimoto and Carlos Guestrin, and his research focuses on developing methods to ensure that machine learning and artificial intelligence pipelines are trustworthy and secure. Specific areas include improving privacy-preserving machine learning and secure computing of machine learning systems.


In academia, he is also an active contributor and has served as a reviewer for many machine learning conferences, winning the Outstanding Reviewer Award at the International Conference on Machine Learning (ICML) in 2022.

From May to August 2017, he worked as a student researcher at the University of Toronto. From January 2018 to July 2019, he worked as a student researcher at the Vector Institute in Toronto. In the summer of 2018, he interned at Google in Mountain View, California, as a Software Engineering Intern, working on the TensorFlow team. This experience gave him a deep understanding of large-scale machine learning frameworks.

In July 2019, he officially joined Google as an artificial intelligence resident researcher, and later became a student researcher, working at Google for a total of 1 year and 9 months. It is worth noting thatThe team he is part of is the (former brain) team led by Geoffrey Hinton, one of the founders of the deep learning field.This experience undoubtedly gilded his research career.

In the summer of 2022, he conducted a 4-month research internship at Microsoft's Redmond headquarters, focusing on differential privacy machine learning. According to him, the results of this research have been deployed in products including Outlook, saving Microsoft considerable money while protecting user privacy.

In February 2024, Xuechen Li joined Elon Musk's newly established artificial intelligence company xAI as a member of the technical team. He is one of the company's earliest 20 employees and is directly involved in the development and training of the Grok artificial intelligence model, xAI's core product. This job puts him at the forefront of the development of artificial intelligence technology and gives him access to the company's core and most sensitive technical information. His tenure at xAI is scheduled to last until July 2025.

However, things didn't go as planned.

02

The lawsuit filed by xAI details the series of events that led to the lawsuit. According to the complaint, the incident occurred in the summer of 2025. In June of that year, Xuechen Li sold $4.7 million worth of xAI stock. He then sought more liquidity and sold an additional $2 million worth of stock to xAI in July. Other reports say he persuaded xAI to repurchase $2.2 million worth of his stock in total, which, combined with the $4.7 million he received earlier in the summer, brings the total cash out to nearly $7 million.

The lawsuit alleges that on July 25, 2025, the same day that Xuechen Li received the last cash proceeds from the stock sale,He copied the company's "confidential information" and "trade secrets" from his company-issued work laptop to "a non-xAI physical or online storage system that he personally controlled."xAI said that Xuechen Li took a variety of measures to cover up his actions, including deleting browser history and system logs, renaming files, and compressing files before uploading to personal devices.

Three days later, on July 28, Xuechen Li resigned from xAI. At this time, he has accepted a job invitation from competitor OpenAI and plans to officially join the company on August 19.

xAI said it discovered Xuechen Li’s behavior during a routine log review on August 11. The unusual activity was recorded by security software deployed by the company to detect and prevent data breaches. After discovering the situation, xAI contacted Xuechen Li via email on the same day and asked him to return and delete all stolen data. It is said that Xuechen Li hired a criminal lawyer to handle the matter at this time.

On August 14, at a meeting between xAI’s lawyers and Xuechen Li’s legal team, things further developed.xAI alleges in its complaint that Xuechen Li admitted orally at the meeting and in a handwritten document that he knowingly misappropriated xAI’s confidential information and attempted to cover up his theft.

He gave xAI permission to inspect a copy of his personal laptop. However, xAI claims that he refused to provide passwords to some key accounts that could reveal the full scope of the theft and the final destination of the data.

Musk himself also weighed in on the matter. He posted on the social platform Musk wrote: "He accepted an offer from OpenAI and then uploaded our entire code base!" This accusation raised the seriousness of the incident to a new level.


An artificial intelligence company’s code base is a reflection of its core values. If Musk’s statement is true, this means that Xuechen Li may have mastered almost all the technical achievements of xAI since its establishment.

03

xAI emphasized the value of the stolen information in lawsuit documents.The company claimed that the stolen trade secrets contained "cutting-edge AI technology with capabilities superior to those offered by ChatGPT and other competing products."If this information is obtained by competitors, it may "save billions of dollars in R&D funds and years of engineering efforts" for the other party, thereby gaining a potential advantage in the fierce competition in the artificial intelligence market.

In the context of the current market value of artificial intelligence model companies reaching hundreds of billions or even trillions of dollars, protecting their core code and model weights is undoubtedly one of the company's highest priorities. It has become normal for researchers to move between different laboratories. How to prevent the loss of core technologies has become one of the biggest survival risks faced by these companies.

Based on the above accusations, xAI made multiple requests to the court:

First, the company asked the court to issue a temporary restraining order forcing Xuechen Li to give up access to any personal devices or online storage services where confidential information may be stored, and to require him to return all confidential materials to the company.

Second, xAI hopes the court will temporarily ban Xuechen Li from working at OpenAI or any other competing company until the company confirms that all its trade secrets have been restored.

Third, xAI also requested the court to order Xuechen Li to compensate for its economic losses, but the specific amount was not specified in the complaint. One thing that is clear is that the amount of compensation xAI requires from Xuechen Li will be very large.

It should be noted that OpenAI is not a defendant in this lawsuit. However, this incident undoubtedly places OpenAI in a delicate position. xAI stated in the lawsuit that OpenAI did not instruct Xuechen Li to steal xAI’s secrets. On the contrary, Xuechen Li used this as a pledge to join OpenAI.

So a question arises here. Judging from Xuechen Li's personal resume, he does not need to use such means to join OpenAI. On the contrary, a simple and ordinary professional interview is far more efficient than going to great lengths to steal xAI's business secrets.

Therefore, xAI and some netizens believe that based on Xuechen Li’s development experience, he may be suspected of falsifying his resume. At the same time, some people point out that OpenAI is not Xuechen Li’s destination, and Google, Microsoft, and Meta may also have business secrets stolen by Xuechen Li.

So did Xuechen Li really fake it?

The picture is the submission record of tatsu-lab, the alpaca large model group of Stanford University on Github, from October 2023 to December 2023. The submitter is Xuechen Li. Phi-2 was a small language model that attracted much attention at the time (end of 2023). SFT (Supervised Fine-Tuning) and DPO (Direct Preference Optimization) are key technologies for fine-tuning and aligning models.

Not only that, in terms of submission, Xuechen Li split a large function into multiple small, independent functions for submission. This is a very good coding habit. Overseas, this approach is called atomic commits.

Xuechen Li's research keeps up with hot topics in the field and uses atomic submission. From a developer's perspective, he does not look like someone who is falsifying his academic qualifications.


The lawsuit comes against the backdrop of an intensifying “war for talent” in the artificial intelligence industry. This chaotic situation is not surprising. Xuechen Li's resignation from xAI and his plan to join OpenAI is itself a microcosm of this war for talent.

The lawsuit also comes amid ongoing tensions between Musk and OpenAI. In recent years, he has repeatedly publicly criticized OpenAI for straying from its original non-profit mission of benefiting all mankind. He has separately sued OpenAI and its CEO Sam Altman, accusing it of violating its original intentions.

OpenAI countersued Musk for harassment in April 2025. Additionally, in August 2025, Musk’s xAI filed another lawsuit against OpenAI and Apple in Texas, accusing the two companies of colluding to monopolize the market for artificial intelligence chatbots on Apple devices. Therefore, xAI’s lawsuit against Xuechen Li can be seen as another new battlefield in a series of conflicts between Musk and OpenAI.

The outcome of this case will have a profound impact on the artificial intelligence industry. It is not only related to Xuechen Li's personal career and reputation, but also to the core technology security of xAI company. More importantly, it will set a new precedent for talent mobility and intellectual property protection within the industry.

How the court defines the scope of trade secrets, how to assess potential losses, and how to balance individual freedom of employment with the company's right to protect its core assets will all become the focus of the industry. Regardless of the final verdict, this lawsuit has made it clear that in the high-investment, high-return, and high-risk field of artificial intelligence, competition around technology, talent, and business interests has extended from laboratories and server rooms to the courts.