SpaceX IPO Filing Confirms xAI Safety Risks Flagged by Watchdog Letter

Former OpenAI employees and AI safety nonprofits published a letter on May 19, 2026, warning that xAI's safety failures could jeopardize SpaceX's planned initial public offering. The letter, coauthored by Guidelight AI Standards cofounders Page Hedley and Steven Adler, argued that xAI lacks industry-standard safety frameworks compared to peers like Google DeepMind and Anthropic, creating unpriced regulatory and litigation risks for SpaceX's effort to raise up to $75 billion. The signatories cited Grok's generation of sexualized images of women and children and its discussion of white genocide, which drew scrutiny from 37 US attorneys general. They demanded SpaceX disclose whether xAI will continue developing frontier models and publish a public safety and governance plan.

Within days, SpaceX's own pre-IPO S-1 filing substantiated those concerns. Submitted on May 20, the document warned potential investors that Grok poses significant reputational and legal risks. SpaceX disclosed that Grok's NSFW modes are more irreverent and harsher than standard offerings and can generate explicit, nonconsensual, or exploitative imagery, including sexualized content involving children. The company is facing multiple lawsuits and investigations in the United States, an inquiry from the Irish Data Protection Commission regarding processing of EU citizens' personal data, and a Federal Trade Commission inquiry into AI chatbot safety for children and teens. SpaceX stated it intends to defend itself vigorously. Elon Musk, who previously pushed for spicier AI output, has denied awareness of Grok generating naked underage images.