Former OpenAI Staffers Warn xAI’s Poor Safety Record Could Complicate SpaceX’s IPO
Two former OpenAI employees and a group of AI security nonprofits are warning that Elon Musk’s AI lab, xAI, could become a liability for potential investors in SpaceX, which is preparing to file what is expected to be the largest initial public offering in Wall Street history.
In a letter to investors released Tuesday, the former staffers highlighted what they described as “unassessed risks” related to xAI that could complicate SpaceX’s announced plans to raise up to $75 billion in its IPO. The rocket company’s private valuation soared to more than $1 trillion after acquiring xAI last year. Musk has claimed his rocket company could launch data centers into space for its AI lab, but the letter’s authors say xAI’s poor security record could complicate how investors view the combined company as it prepares to submit its IPO prospectus filing.
One of the letter’s signatories and co-authors is a new nonprofit called Guidelight AI Standards, co-founded by former OpenAI security researcher Steven Adler and former OpenAI policy advisor Page Hedley. The group, supported by private donors, aims to improve the security practices of pioneering AI companies. Other nonprofits related to AI safety have also signed up, including Legal Advocates for Safe Science and Technology, Encode AI, and The Midas Project.
Hedley told WIRED in an interview that he believes xAI has the worst security practices “almost across the board” compared to other leading AI developers, including OpenAI, Google DeepMind and Anthropic. As a result, he argues, SpaceX could face greater regulatory and litigation risk than other AI labs.
The letter’s authors argue that SpaceX should disclose several pieces of information to investors, including whether xAI intends to continue developing pioneering AI models. SpaceX recently struck a deal to sell a significant portion of its GPU capacity to Anthropic, and the letter claims the deal “does not make it clear whether xAI is still a frontier AI competitor within a larger holding company.” If xAI continues to develop frontier AI models, the authors say it should be required to publish a public safety and governance plan.
SpaceX and xAI did not immediately respond to WIRED’s request for comment.
The letter also outlines examples of how xAI failed to follow industry standard security practices, such as publishing detailed frameworks to mitigate risks related to the use of its AI models in cyberattacks. The authors also describe specific security incidents at xAI that they believe warrant further investigation. Among the most notable were Grok, xAI’s flagship AI chatbot, which spontaneously brought up white genocide in its responses. In another case, xAI allowed Grok to generate thousands of sexualized images of women and children, which spread widely on Musk’s X social media platform. This latest case prompted at least 37 US attorneys general to send a letter demanding that Musk’s AI lab take steps to protect women and children on its platform.
Hedley says the number of security incidents xAI has encountered and the regulatory attention they have received are “far out of proportion to its market share.” As lawmakers grow increasingly alarmed by the cyber capabilities of advanced AI models like Anthropic’s Claude Mythos, new security regulations could be on the horizon. The Trump administration is already reportedly considering an executive order that would give U.S. intelligence agencies more control over AI models.
“It takes serious investments to reign [AI safety] risks, and it appears that xAI has historically underinvested here,” Adler says. The letter cites a Washington Post article that said xAI had only “two or three” people working on security as of January. “One question investors should ask is: If xAI stays on the frontier, how much might it cost, in fact, to manage these risks. [risks] responsibly? If they don’t, what could be the consequences?
