.jpg?w=1024)
Elon Musk made a rare public U-turn, saying he underestimated Anthropic and now believes the AI startup sits at the top of the industry.
After months of publicly criticizing Anthropic, Musk acknowledged on X that his earlier assessment of the company was incorrect.
"I was clearly wrong about Anthropic. They are obviously currently the leader in AI. No company has released a model as good as Mythos/Fable and they will undoubtedly have Mythos 2 ready soon," Musk wrote.
The statement marks a dramatic change in tone from comments he made over the past year, when he repeatedly questioned Anthropic's direction and chances of competing with the biggest AI developers. Among other remarks, Musk previously argued that "winning was never in the set of possible outcomes for Anthropic."
His latest comments came in response to an X user who claimed Musk could cripple Anthropic because the company relies on computing infrastructure provided by SpaceXAI.
Musk rejects using compute as leverage
Rather than embracing the suggestion, Musk said he would not use that business relationship to harm a competitor.
"That’s not my style," he wrote. To support that point, Musk pointed to decisions he says demonstrate his approach to competition.
"Tesla open sourced its patents, and we made the Supercharger network available to all competitors, even though we could have made it a walled garden," he wrote. "SpaceX launches competing satellite systems with no increase in price or use of unfair terms. [...] Even my worst enemies can attack me on this platform."
Partnership changed the relationship
The comments come just weeks after SpaceX and Anthropic entered a major infrastructure partnership.
According to SpaceX’s S-1 filing, anthropic secured access to more than 300 megawatts of compute capacity at SpaceX's Colossus 1 data center, representing more than 220,000 Nvidia GPUs. SpaceX's S-1 filing also showed Anthropic agreed to pay roughly $1.25 billion a month for the service through May 2029, although either company can end the agreement with 90 days' notice.
Since announcing the partnership in May, Musk's tone toward Anthropic has steadily softened. He previously said he spent time with the company's senior leadership and came away impressed, writing: "Everyone I met was highly competent and cared a great deal about doing the right thing. No one set off my evil detector."
Even so, Musk continued to frame Anthropic as a rival, noting earlier this year that SpaceXAI is much younger than both Anthropic and OpenAI and suggesting the competitive landscape could change over time.
What the reversal signals for the AI race
Musk's comments stand out because they come from the head of a competing AI company that recently launched its own frontier models while simultaneously supplying critical infrastructure to a direct rival.
His acknowledgment also highlights how quickly leadership can shift in generative AI, where new model releases can reshape perceptions within weeks.
At the same time, Anthropic's dependence on leased compute reflects a growing reality in the industry: even leading AI developers often rely on competitors or external partners for the enormous computing power required to train and deploy advanced models.
That combination of fierce competition and deep commercial partnerships is becoming increasingly common as AI companies race to build more capable systems while sharing parts of the underlying infrastructure.
For businesses and enterprise customers, it also shows that leadership in AI is no longer defined solely by model performance, but also by access to compute, strategic partnerships, and the ability to scale rapidly without disrupting customers.
Also read: OpenAI and Google’s reported AI access for Chinese firms through Singapore shows how frontier AI competition is increasingly shaped by infrastructure, restrictions, and cross-border demand.