Elon Musk’s company could grant the US Department of Defense access to its computing power, a move that threatens to deepen the US military’s reliance on a single contractor.
The company already works closely with the military, providing rocket launches, satellite communications, and missile-tracking services. If the new agreement is signed, the Pentagon will also rely on Musk’s infrastructure for AI. Critics note that such a deep reliance on a single businessman’s services raises serious concerns. Sources close to the matter specify that discussions are still ongoing and could ultimately fall through.
However, SpaceX’s ambitions go beyond government contracts. The company plans to compete directly with cloud providers like CoreWeave by selling computing power to AI firms at lower prices.
This is not the company’s first move to secure its footing in the computing infrastructure market. Earlier in June, SpaceX struck a $30 billion cloud computing deal with Google. Under the agreement, the tech giant will pay $920 million per month from October 2026 through June 2029 for access to Nvidia chips and related AI capacity.
Anthropic has also agreed to pay SpaceX $1.25 billion per month through May 2029. This money will cover computing resources at the Colossus data centers in Memphis, Tennessee. Anthropic will gain access to more than 300 megawatts of capacity, which its representatives say will directly improve performance for paid Claude Pro and Claude Max subscribers.
Another major client is Reflection AI. Starting July 1st, 2026, the startup will pay roughly $150 million per month to access Nvidia GB300 chips at the Colossus 2 data center. If the contract runs through 2029, its total value could reach $6.3 billion.
SpaceX is gradually turning into one of the key players in the artificial intelligence market. A successful deal with the Pentagon could be the decisive step that firmly establishes the company as a global provider of computing infrastructure.
