Elon Musk’s AI Startup Claims OpenAI Lured Engineers to Misappropriate Confidential Code and Training Methods

Elon Musk’s xAI Accuses OpenAI Of Stealing Trade Secrets Through Employee Poaching

Elon Musk’s artificial intelligence company xAI has taken OpenAI to California federal court, alleging that its rival orchestrated a “coordinated, unfair, and unlawful campaign” to obtain proprietary technology.

The complaint, filed on 24 September, claims OpenAI aggressively recruited xAI engineers and encouraged them to misappropriate confidential material, including source code, training techniques, and strategies for data centre deployment.

From Co-Founder To Fierce Rival

Musk co-founded OpenAI in 2015 alongside Sam Altman, Greg Brockman, and Ilya Sutskever, but left its board in 2018 citing conflicts with Tesla’s work on self-driving cars.

Since then, he has frequently clashed with the company, filing multiple lawsuits.

The latest case comes just weeks after Musk’s firms launched an antitrust action against Apple and OpenAI over their iPhone integration deal.

Engineers At The Centre Of The Dispute

The filing names several former xAI staff.

Early engineer Xuechen Li allegedly uploaded the entire xAI source code base to a personal cloud account in July 2025, later confessing in writing to misappropriating company materials.

Within hours of his encrypted Signal exchanges with OpenAI recruiter Tifa Chen, the lawsuit claims, Li copied files before receiving a multi-million-dollar job offer.

Another engineer, Jimmy Fraiture, is accused of using Apple’s AirDrop feature to transfer confidential code “at least five times” after joining OpenAI.

The lawsuit also cites an unnamed senior finance executive, later identified as Mike Liberatore, xAI's former Chief Financial Officer.

The executive reportedly praised xAI’s data centre speed as its “secret sauce” before leaving for OpenAI, where he took up a role in data centre spending despite lacking prior AI experience.

@xAI innovates; OpenAI cheats.

Read the full complaint against OpenAI:https://t.co/bLLSHyuRR4https://t.co/SMvC5rJ9cZpic.twitter.com/hlGsiKyzwH
— Lily Lim (@LiLiDuc22) September 25, 2025
a

Recruiter’s Role Under Scrutiny

The complaint places particular focus on OpenAI recruiter Tifa Chen, who allegedly targeted multiple xAI employees at once.

According to the filing, Chen reacted with the phrase “no way!” after Li copied source code, a detail that xAI says links the recruiter directly to the misappropriation.

Whether recruiters acted with the company’s knowledge could become a key legal test.

Legal Experts Weigh The Case

Lawyers following the dispute say the case hinges on whether poaching employees crossed into theft of trade secrets.

Navodaya Singh Rajpurohit of Coinque Consulting observed that “hiring alone is rarely enough to prove trade-secret misuse.”

Ishita Sharma of Fathom Legal added that xAI must carefully define its “secret sauce,” grouping together elements such as GPU racking, vendor contracts, and orchestration playbooks without revealing technical diagrams in open court.

For OpenAI’s defence, Sharma suggested the company could present “time-stamped records: internal Git commits, R&D notes, supplier invoices, and emails” to prove independent development, noting that earlier records would be the most persuasive.

What xAI Wants From The Court

xAI is demanding damages, restitution, and injunctions requiring OpenAI to purge all materials allegedly taken from its systems and to destroy any models trained with that information.

Neither company has yet issued public comment on the lawsuit.

The High-Stakes Battle Over AI Talent And Trade Secrets

From Coinlive’s perspective, the case shines a spotlight on the brutal talent war driving Silicon Valley’s AI race.

Musk’s xAI positions itself as a challenger to OpenAI’s dominance, yet its survival may hinge on whether it can protect its intellectual property long enough to scale.

While poaching talent is nothing new, the line between recruitment and theft is now being tested in court.

If xAI succeeds, it may slow OpenAI’s momentum.

If it fails, in the current AI market, keeping your brightest minds loyal may be the only real defence.