SpaceX's AI Ambitions and xAI's Financial Burn
SpaceX's recent IPO filing has laid bare the company's substantial investment in artificial intelligence, primarily through its xAI division, revealing a significant financial burn rate. In 2025, xAI reportedly burned through $6.4 billion, with capital expenditures for the AI segment reaching $12.7 billion, far exceeding spending in the Space and Connectivity businesses. For the first quarter of 2026 alone, the AI division accounted for $7.72 billion of SpaceX's total capital expenditures of $10.1 billion, and posted a loss of nearly $2.5 billion. This aggressive spending is directed towards developing data centers, acquiring GPUs, and attracting engineering talent to compete with established AI leaders like OpenAI and Anthropic.
Despite these substantial losses, SpaceX views AI as a critical component of its future, outlining a potential $26.5 trillion market opportunity linked to artificial intelligence within a projected $28.5 trillion total addressable market across its businesses. The company's strategy involves building orbital data centers and AI-powered systems, with the ultimate goal of supporting a multi-planetary economy. The filing underscores that while Starlink remains a profitable engine for SpaceX, generating $11.4 billion in revenue and $4.4 billion in operating income in 2025, the AI segment is currently a significant drain on resources.
The Anthropic Deal: A Glimmer of AI Revenue
Amidst xAI's considerable expenditures, SpaceX's IPO filing revealed a significant revenue stream through a compute capacity agreement with AI developer Anthropic. Anthropic has committed to paying xAI $1.25 billion per month through May 2029 for access to large-scale AI compute capacity. This agreement grants Anthropic access to 300 megawatts of compute capacity from the Colossus 1 data center near Memphis, Tennessee, and will also include capacity from the newer Colossus 2 data center.
The contract, which could generate over $40 billion in total revenue for SpaceX if it runs for the full term, allows either party to terminate with 90 days' notice. This deal highlights SpaceX's strategy to monetize its substantial AI infrastructure, with Elon Musk indicating that the company is in discussions with other entities for similar compute-as-a-service agreements. This partnership places two major AI competitors in a commercial relationship, offering Anthropic much-needed resources for its growing AI models and providing SpaceX a crucial revenue channel for its AI segment.
Elon Musk: The Central Figure and Risk Factor
The SpaceX IPO filing unequivocally positions Elon Musk at the center of the company's operations and future, simultaneously identifying him as a significant risk factor for investors. Musk holds approximately 42.5% of SpaceX equity but commands 83.8% of voting power through Class B super-voting shares, effectively giving him outright control over his removal from his roles. This unusual governance structure, which Harvard Law professor Lucian Bebchuk described as "not common," strips public investors of the right to remove the chief executive.
The S-1 filing explicitly flags Musk's potential loss as a multi-page risk factor, citing his overlapping commitments at Tesla, xAI, X, Neuralink, and The Boring Company. Analysts have warned of an "Elon Musk effect," predicting that SpaceX's stock could experience substantial volatility tied to Musk's actions and public statements, similar to Tesla's stock history. Furthermore, Musk's compensation package is tied to audacious targets, including SpaceX reaching a $7.5 trillion market capitalization and establishing a permanent human colony on Mars.
Grok's "Spicy" Mode and Associated Risks
SpaceX's IPO filing also addresses the inherent risks associated with xAI's chatbot, Grok, particularly its "spicy" mode. The company explicitly warned investors about the potential for Grok to generate "nonconsensual or exploitative imagery" and content that could be deemed "harmful, harassing, abusive, or discriminatory." The S-1 paperwork cited ongoing investigations into allegations that Grok created "content representing children in sexualized contexts" and referenced multiple lawsuits stemming from Grok's image-generation and editing features.
SpaceX has set aside over $500 million for potential litigation losses related to these allegations, underscoring the serious financial and reputational implications. The "spicy" mode, designed to offer more unfiltered interactions, has drawn significant regulatory scrutiny in various countries, including the US, Ireland, Canada, the UK, and Brazil. This highlights the challenges and liabilities associated with developing and deploying advanced AI models, especially those designed with fewer content restrictions.