FOLLOW US

Elon Musk’s artificial intelligence venture xAI urged a California federal judge Tuesday to reconsider her tentative decision to dismiss its lawsuit accusing OpenAI of orchestrating a worker exodus to siphon off trade secrets.

At a tense hearing, xAI’s counsel argued that while individual allegations against former employees may appear fragmented, taken together they form a compelling picture — what she repeatedly called the “whole gestalt” — sufficient to sustain claims against the ChatGPT maker.

Judge Signals Skepticism

The dispute centers on the XAI OpenAI IP Theft Suit, filed in September by xAI Corp., the company behind the large language model Grok. The lawsuit alleges OpenAI embarked on a targeted hiring campaign aimed at xAI employees with access to sensitive infrastructure and proprietary code.

Those workers, according to the complaint, were involved in building xAI’s data centers and developing Grok.

Ahead of Tuesday’s hearing, U.S. District Judge Rita F. Lin indicated in a notice that she was inclined to grant OpenAI’s motion to dismiss, writing that xAI appeared to have fallen short of alleging sufficient facts showing OpenAI poached employees or misappropriated source code.

“The Whole Gestalt”

LeElle Slifer of Winston & Strawn LLP, representing xAI, acknowledged that without discovery the company cannot flesh out more granular details. But she pressed the court to evaluate the allegations collectively rather than in isolation.

“I would encourage Your Honor to not separate each of these facts, but rather look at the facts together,” Slifer said.

She argued that multiple employees left within a compressed timeframe and were allegedly involved in taking source code and confidential materials. Viewed as a pattern, she said, the claims plausibly suggest that OpenAI induced trade secret theft.

“If this was just one or two workers it would be one thing,” Slifer said. “But this is several different employees during a very short period of time all unquestionably stealing source code and trade secrets.”

Signal App and Source Code at Issue

To bolster the narrative, Slifer pointed to communications through the encrypted messaging app Signal between OpenAI recruiters and departing xAI employees. She also cited allegations that at least one worker downloaded source code onto a personal device, while others retained confidential materials during interviews with OpenAI.

Slifer argued that the use of Signal — which does not retain stored communications — supported an inference of secrecy.

Judge Lin, however, questioned whether such facts are enough to infer that OpenAI directed employees to take intellectual property. She also appeared unconvinced that the mere use of Signal amounted to a “huge red flag.”

The judge further noted that the alleged methods of taking xAI’s intellectual property differed from employee to employee and “seem to have nothing to do with one another,” casting doubt on whether the pleadings support an inference of coordinated direction by OpenAI.

Slifer countered that a common thread ties the workers together: the same recruiter allegedly communicated with several of them.