Two former OpenAI employees, along with a coalition of nonprofit organizations focused on AI safety, have issued a public warning to potential investors in SpaceX. They express concerns that xAI, an AI lab owned by Elon Musk, could represent a significant financial and reputational liability for the aerospace giant.
In an open letter, the authors highlight numerous "unaccounted risks" associated with xAI's operations. These risks could complicate SpaceX's ambitions to raise up to $75 billion in an upcoming IPO. Following the acquisition of xAI last year, the private valuation of the rocket manufacturing company surged past $1 trillion. Musk previously claimed that SpaceX would be able to launch massive data centers into space to support his AI lab. However, the letter's authors emphasize that systematic failures in xAI's safety could tarnish investor sentiment toward the integrated entity.
One key author and signatory of the document is Guidelight AI Standards, a newly established nonprofit led by former OpenAI security researcher Stephen Adler and ex-OpenAI policy advisor Paige Headley. Funded by private donors, this organization aims to enhance safety standards among leading AI developers. The letter has also garnered support from other relevant organizations, such as Legal Advocates for Safe Science and Technology, Encode AI, and The Midas Project.
In an interview, Headley criticized xAI for displaying some of the worst safety practices among leading AI developers, including OpenAI, Google DeepMind, and Anthropic. As a result, the combined entity of SpaceX could face severe regulatory pressure and extensive legal challenges.
The letter cites various instances where xAI has disregarded industry-standard safety protocols. The lab has yet to release foundational frameworks to prevent the misuse of its models in cyberattacks. Notable incidents include the chatbot Grok generating racist content regarding "white genocide" and enabling users to create and distribute thousands of sexualized images of women and children across social media platforms. This latter incident prompted official demands from attorneys general in at least 37 U.S. states for Musk to protect minors on his platform.
Headley noted that the frequency and severity of scandals surrounding xAI are vastly disproportionate to its actual market share. With growing concerns from lawmakers over the cyber capabilities of advanced models like Anthropic's Claude Mythos, new safety regulations could be imminent. Reports indicate that the Trump administration is preparing an executive order to expand oversight of AI labs by U.S. intelligence agencies.
"Managing AI risks requires substantial funding, and xAI has historically underinvested in this area," Adler stated. The letter references an investigation revealing that as of January, only two or three individuals at xAI were dedicated to safety issues. Investors should contemplate the costs of responsibly managing these risks if xAI intends to remain at the technological forefront, as well as the potential consequences of ignoring them.
While the authors acknowledge recent improvements in Musk's project, such as an expanded agreement with the White House to test new AI models before their official release, they stress that much greater transparency is necessary for Wall Street to adequately assess the risks associated with SpaceX. The former OpenAI employees conclude that xAI's troubling past warrants serious scrutiny, but this alone does not pave the way for resolution. The creators of Guidelight AI Standards aim to establish a set of safety benchmarks that will serve as clear guidelines for policymakers, the press, and investors.
This warning underscores the potential volatility in SpaceX's market strategy, potentially giving competitors an edge if they can address safety concerns more effectively.
Informational material. 18+.
