SpaceX recasts IPO around AI ambitions

SpaceX is pitching its planned public listing as a bet on artificial intelligence as much as rockets, telling prospective investors that AI services could open a $26.5 trillion market opportunity and transform the company from a launch-and-satellite operator into a global computing platform.

The shift marks a striking turn for Elon Musk’s space company as it prepares for what could become the largest stock-market debut on record. The offering is being built around three linked businesses: reusable rockets, Starlink satellite internet and an AI operation strengthened by the February purchase of xAI, the developer of the Grok chatbot.

The company is seeking a valuation of about $1.75 trillion, a level that would place it among the world’s most valuable listed businesses from its first day of trading. SpaceX is expected to list on Nasdaq and Nasdaq Texas under the ticker SPCX, with major Wall Street banks leading the sale. A roadshow is targeted for early June, with the listing possible later that month.

The pitch to investors rests on the view that demand for AI models, enterprise automation and large-scale computing will move beyond conventional data centres. SpaceX says its long-term plans include AI infrastructure on Earth and, eventually, space-based data centres powered by solar energy. That ambition is meant to set it apart from OpenAI, Anthropic, Google, Microsoft and Amazon, all of which are competing for customers, chips, power supplies and cloud capacity.

Financial disclosures show the scale of the bet. SpaceX reported first-quarter revenue of $4.69 billion and an operating loss of $1.94 billion. Starlink generated an operating profit of $1.19 billion, but the AI division recorded a $2.47 billion loss on $818 million in revenue. AI also accounted for 76% of the company’s $10.1 billion in capital expenditure during the quarter, underscoring how quickly spending has shifted towards computing infrastructure.

A major commercial validation came through a compute agreement with Anthropic, under which the Claude developer is to pay $1.25 billion a month through May 2029 for access to SpaceX’s Colossus and Colossus II data-centre clusters in Memphis, Tennessee. The agreement could generate $15 billion a year, although it includes a 90-day termination clause and lower fees during the capacity ramp-up period.

SpaceX’s defenders argue that the company has repeatedly turned improbable engineering goals into profitable platforms. Reusable Falcon rockets cut launch costs, Starlink created a consumer and enterprise broadband business from thousands of low-Earth-orbit satellites, and Starship is intended to support heavier payloads, lunar missions and Mars plans. The same logic is now being applied to AI: combine cheap launch capacity, satellite networks, energy access and model development to build an infrastructure stack few rivals can match.

The risks are equally prominent. Grok remains behind OpenAI, Google and Anthropic in many enterprise and government settings, despite aggressive pricing. Federal AI inventory data showed only limited xAI or Grok use across publicly listed civilian government applications, while rival systems appeared far more widely. Corporate usage data also suggested Grok has struggled to gain meaningful traction among business users, raising questions over whether SpaceX can capture the market share implied by its valuation.

Regulatory and legal exposure has also entered the investor debate. SpaceX has disclosed lawsuits tied to Grok’s image-generation and editing features, including claims linked to non-consensual explicit imagery and content involving children in sexualised contexts. The company has said it will defend itself vigorously, while also warning investors that AI products may create reputational, legal and regulatory risks.

Governance is another concern for public-market investors. SpaceX plans a dual-class structure giving Class B shareholders 10 votes per share, while Class A shares sold to the public will carry one vote each. Musk is expected to retain dominant voting control, and the company’s governance provisions could limit shareholder influence over strategy, disputes and leadership.

The IPO therefore offers investors an unusually broad proposition: a profitable satellite-internet arm, a capital-heavy space programme, a loss-making AI business, and a founder whose reputation remains central to both the company’s premium and its risk. Demand for AI compute is rising sharply, but the industry is also constrained by power availability, chips, water use, local opposition to data centres and fast-changing model economics.