SpaceX's $45 Billion AI Deal with Anthropic May End Early, Says Musk

SpaceX might lose out on billions of dollars from a major artificial intelligence deal much sooner than expected, according to CEO Elon Musk.

The space company signed a deal to rent out its powerful computers to Anthropic (an AI company that makes ChatGPT competitor Claude). Under the agreement, Anthropic would pay SpaceX $1.25 billion per month through May 2029 to use SpaceX's data centers (massive buildings filled with computers) in Tennessee.

If the deal lasted the full three years, SpaceX could earn up to $45 billion. However, Musk revealed on X (formerly Twitter) that the deal is actually much shorter:

• The initial agreement is only for 180 days • Either company can cancel with 90 days' notice after that • SpaceX specifically requested the short-term arrangement

Why does this matter? The deal was seen as a way for SpaceX to make money from its expensive AI infrastructure (computer systems) while its own AI product, Grok, struggles to attract users. Only 21% of SpaceX users have tried Grok's AI features.

SpaceX's AI division lost $2.5 billion in the first quarter of 2026 while earning just $818 million in revenue (money coming in). The company is betting big on AI, believing it will eventually become its largest business.

The shorter deal timeline suggests SpaceX might be less confident about keeping Anthropic as a long-term customer, potentially costing the company billions in future revenue.

This is an AI-generated summary. Read the original article at: https://www.marketwatch.com/story/spacexs-major-ai-compute-deal-with-anthropic-could-end-a-lot-sooner-than-expected-30f46b83?mod=mw_rss_topstories