Yesterday's enemy becomes today's vendor, principle yields to opportunity.
Three months after calling Anthropic 'evil,' Elon Musk has agreed to a $4 billion arrangement making SpaceX a provider of computational infrastructure to the very company he condemned. The reversal, announced in May 2026, invites the perennial question that follows powerful figures through history: whether public moral pronouncements are convictions or simply positions held until the price of changing them becomes attractive enough. In the accelerating contest over artificial intelligence, it appears that yesterday's adversary can become today's dependent tenant, and that the architecture of influence is often built not through argument, but through control of essential resources.
- A $4 billion deal between SpaceX and Anthropic has exposed the gap between Musk's public condemnations and his private commercial calculations.
- The arrangement grants Musk structural leverage over Anthropic's operations — he can cut off data supply if the company deploys its AI tools in ways he deems unacceptable.
- Anthropic, pressed by competition from OpenAI and other well-funded rivals, accepted the terms as a trade between independence and the computational power needed to stay in the race.
- The deal has unsettled observers who saw Musk as a principled critic of AI consolidation, now watching him become an infrastructure landlord shaping the field from within.
- Whether the partnership holds remains uncertain — Musk's history of abrupt reversals and public denunciations of former allies leaves the arrangement's durability genuinely in question.
Three months ago, Elon Musk called Anthropic evil. The word was not casual — it came from a man who had left OpenAI in conflict, built xAI as a direct rival, and cast himself as a watchdog against AI's misuse. Then, in May 2026, he announced a $4 billion deal making SpaceX a data infrastructure provider to the same company he had condemned.
The arrangement came with explicit conditions. Musk made clear to Anthropic's leadership that SpaceX would retain leverage over how the company deployed its AI tools — a threat to cut off computational resources if his terms were violated. It was control dressed as partnership, a way to shape Anthropic's trajectory without owning it outright.
For Anthropic, the choice was pragmatic: the company needed infrastructure to compete, and SpaceX could provide it. The cost was accepting Musk's oversight. Capability won over independence.
What made the deal stranger was its contradiction with Musk's public posture. He had warned against unchecked AI development and corporate consolidation, yet here he was embedding himself in one of the field's most prominent companies through the terms of a commercial contract — influence operating below the level of public argument.
Whether the arrangement will hold is another question. Musk's partnerships have a history of rapid deterioration. For now, the deal stands as a monument to the AI race's strange logic: yesterday's enemy becomes today's vendor, and principle yields to the next shift in incentives.
Three months ago, Elon Musk was unsparing in his assessment of Anthropic. The company, he said, was evil. The word carried weight coming from him—a man who had co-founded OpenAI before his acrimonious departure, who had built xAI as a direct competitor to the very firms he now criticized, who saw himself as a guardian against the misuse of artificial intelligence. Evil was not hyperbole in his vocabulary. It was diagnosis.
Then, in May of 2026, Musk announced he would accept a $4 billion deal with Anthropic. The arrangement would make SpaceX, his aerospace company, a data landlord to the AI firm—a provider of computational infrastructure and resources that Anthropic would depend on to train and operate its models. The reversal was not subtle. It was the kind of about-face that invites immediate scrutiny, the kind that makes observers wonder what had changed, or whether the earlier condemnation had ever been sincere.
The deal came with conditions. According to reporting, Musk made his terms explicit to Anthropic's leadership: SpaceX would maintain leverage over how Anthropic deployed its AI tools. If the company violated those terms, if it used its systems in ways Musk deemed unacceptable, the data supply could be cut off. It was a form of control dressed as partnership—a way to remain invested in Anthropic's trajectory without owning the company outright.
The timing raised questions about what had shifted in the intervening months. Had Musk's concerns about Anthropic's ethics evaporated? Had the company made changes he found acceptable? Or had the calculus simply changed: the opportunity to profit from Anthropic's growth, to maintain a foothold in the AI race through infrastructure provision, had outweighed whatever moral objections he had voiced before. The deal suggested that in the high-stakes world of artificial intelligence development, business interests could move faster than principle.
For Anthropic, the arrangement represented a pragmatic choice. The company needed computational resources to compete with OpenAI and other well-funded rivals. SpaceX offered access to significant data infrastructure. The cost was accepting Musk's oversight, his ability to threaten the partnership if Anthropic's decisions displeased him. It was a trade-off between independence and capability, and Anthropic had chosen capability.
The broader context made the deal stranger still. Musk had positioned himself as a critic of the AI industry's direction, a voice warning against unchecked development and corporate consolidation. Yet here he was, becoming a stakeholder in one of the field's most prominent companies, using his control over resources as a lever to shape its behavior. It was a form of influence that operated below the level of public debate, conducted through the terms of a commercial contract rather than through argument or persuasion.
What remained unclear was whether this arrangement would hold. Musk's relationships in tech had a history of deteriorating rapidly. His threats to cancel deals, to withdraw support, to publicly denounce partners—these were familiar patterns. Whether SpaceX's data infrastructure would remain available to Anthropic, whether Musk would find cause to invoke his leverage, whether the partnership would survive the next controversy or disagreement—all of that remained unwritten. For now, the deal stood as a monument to the strange logic of the AI race: yesterday's enemy becomes today's vendor, principle yields to opportunity, and the future remains hostage to the next shift in incentives.
Notable Quotes
Musk made clear to Anthropic that SpaceX would cancel the data supply deal if the AI company deployed its tools in ways he deemed unacceptable— reporting on the deal terms
The Hearth Conversation Another angle on the story
Three months is a short time to go from calling someone evil to handing them four billion dollars' worth of infrastructure. What actually changed?
That's the question everyone's asking, and the honest answer is we don't know for certain. But the deal structure tells you something: Musk didn't just become a passive investor. He built in leverage. If Anthropic uses its tools in ways he doesn't like, he can cut off the data supply.
So he's not really reversing his position. He's just finding a way to control Anthropic from the outside.
That's one way to read it. Another is that he realized he couldn't stop Anthropic from succeeding, so he might as well profit from it and maintain some influence. In the AI race, you either compete or you find a way to stay in the game.
But doesn't that undermine his whole public stance about being the responsible voice in AI development?
Completely. He's positioned himself as a critic of corporate consolidation and unchecked development. Now he's using control over resources to shape a competitor's behavior. It's influence operating through contracts instead of argument.
Do you think this deal actually holds? Given Musk's history with partnerships?
That's the real uncertainty. His relationships tend to deteriorate fast. Whether SpaceX keeps supplying Anthropic, whether he finds reason to invoke his leverage—that's all still unwritten. For now it's a deal. Whether it survives the next disagreement is another question entirely.