Elon Musk’s artificial intelligence venture xAI burnt through $6.4 billion in 2024, according to financial disclosures buried in SpaceX’s initial public offering documents filed this week. The figure represents the first public confirmation of xAI’s operating losses and underscores the extraordinary capital requirements of competing in frontier AI development.
The disclosure, reported by TechCrunch AI, reveals spending levels that dwarf most technology startups and position xAI amongst the most capital-intensive AI operations globally. The company, founded in July 2023, has been racing to build competing infrastructure to OpenAI, Anthropic, and Google’s DeepMind whilst simultaneously developing its Grok family of large language models.
SpaceX’s filing indicates the spending trajectory shows no signs of moderating. xAI is currently expanding its Memphis-based supercomputing facility, known as Colossus, which already houses 100,000 Nvidia H100 GPUs. Plans call for scaling the cluster to accommodate hundreds of thousands of additional chips, requiring billions in additional capital expenditure over the coming quarters.
The financial details emerged because SpaceX holds a significant stake in xAI, necessitating disclosure of material related-party transactions. SpaceX has provided infrastructure support and computing resources to xAI, creating financial entanglements that securities regulators require be made public during the IPO process.
The $6.4 billion burn rate reflects spending across multiple fronts: procuring tens of thousands of advanced AI accelerators at approximately $30,000 per unit, constructing and operating massive data centres with associated power and cooling infrastructure, recruiting top-tier AI researchers commanding seven-figure compensation packages, and funding the computational costs of training increasingly large models.
xAI has raised substantial capital to fund these operations, including a $6 billion Series B round in May 2024 that valued the company at $24 billion. Investors include Sequoia Capital, Andreessen Horowitz, and Saudi Prince Alwaleed bin Talal. Despite the massive losses, the company’s valuation reflects investor confidence in Musk’s ability to compete with established players and potentially monetise AI capabilities through integration with his other ventures, including X (formerly Twitter) and Tesla.
Business Impact
The disclosure reshapes understanding of competitive dynamics in the AI infrastructure race. Nvidia emerges as a clear beneficiary, with xAI representing a multi-billion-dollar customer for its data centre GPUs. Energy providers and data centre operators serving xAI’s facilities gain revenue visibility, whilst competing AI companies face pressure to match or exceed xAI’s infrastructure investments to remain competitive.
For investors, the figures validate concerns about the capital intensity of foundation model development. Companies pursuing similar strategies—including Anthropic, which reportedly seeks $2 billion annually, and OpenAI, with estimated losses exceeding $5 billion in 2024—face questions about paths to profitability and the sustainability of current spending levels without corresponding revenue growth.
The spending also has implications for Musk’s other ventures. Tesla shareholders have questioned whether xAI’s access to Musk’s attention and potentially shared AI research represents value transfer from the publicly traded automaker. SpaceX’s financial exposure to xAI, now public, may factor into investor assessments of the rocket company’s IPO valuation.
What’s Next
Market observers should monitor whether xAI can translate infrastructure spending into competitive model performance and, critically, revenue generation. The company must demonstrate that Grok can compete technically with GPT-4, Claude, and Gemini whilst building a sustainable business model beyond integration with Musk’s existing platforms.
The broader question is whether the AI industry’s current capital expenditure levels—potentially exceeding $50 billion annually across major players—can generate sufficient economic value to justify the investment. xAI’s spending, now quantified, provides a benchmark for assessing the true cost of competing in frontier AI development.
