'Currently the leader in AI': After years of criticising Anthropic, Elon Musk praises its Mythos/Fable model

Elon Musk has rarely had a kind word to say about Anthropic. Over the years, the SpaceXAI CEO has repeatedly criticised the AI startup, questioned its capabilities and even suggested that it was never going to emerge as a winner in the race for frontier AI.

Now, in an unexpected reversal, Musk has publicly acknowledged that he got it wrong.

Responding to a discussion on X about Anthropic's reliance on SpaceXAI's computing infrastructure, Musk admitted that the company has overtaken its rivals in artificial intelligence.

"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.

More from Tech

The post marks a dramatic shift from his earlier stance.

The conversation began after an X user claimed that Anthropic depends heavily on compute rented from SpaceXAI. The user argued that Musk could, in theory, refuse to renew the company's compute contract and significantly disrupt one of his biggest competitors. The post also referenced one of Musk's older comments from 2025, in which he had confidently declared that "winning was never in the set of possible outcomes for Anthropic."

Instead of defending that prediction, Musk openly conceded that he had misjudged the company.

Public admissions of this kind are relatively rare from Musk, particularly when they involve direct competitors.

His remarks also amount to one of the strongest endorsements Anthropic has received from a rival CEO. By calling the company the current leader in AI, Musk effectively acknowledged the rapid progress Anthropic has made in building frontier models that compete with the industry's biggest names.

However, Musk did not stop at praising the company's technology.

Addressing concerns about Anthropic's dependence on SpaceXAI's infrastructure, he insisted that he would never exploit that relationship for competitive advantage.

"I would never cut them off in a way that hurt them badly, even as a competitor. That's not my style," he wrote.

To support that claim, Musk pointed to Tesla's decision to open-source many of its patents, the company's Supercharger network being made available to rival automakers, and SpaceX launching satellites for competing operators without imposing unfair pricing or restrictive terms.

While the statement may reassure some observers, it also highlights an uncomfortable reality within the AI industry.

Training and running frontier AI models increasingly depends on access to enormous computing resources. If a leading AI company relies on infrastructure controlled by a direct competitor, questions about long-term independence are almost inevitable. Contracts can expire, business priorities can change and competitive dynamics can shift quickly.

That does not mean SpaceXAI intends to use its position against Anthropic. Musk's comments suggest the opposite. Yet the discussion itself underscores why AI companies are racing to secure their own computing capacity rather than depending on rivals.

For Anthropic, the biggest takeaway is arguably not Musk's promise, but his recognition. A company he once dismissed is now, by his own admission, setting the pace in AI.

Whether that praise marks the beginning of a more respectful relationship or simply reflects the realities of an increasingly competitive AI race remains to be seen. But one thing is clear: Musk's view of Anthropic has changed dramatically, and that, in itself, is headline-worthy.