For years, Elon Musk has rarely held back when talking about rivals in the artificial intelligence industry. Few companies attracted more criticism from him than Anthropic.
He questioned its business practices, accused it of political bias, dismissed its chances of winning the AI race, and repeatedly argued that the company was headed in the wrong direction.
Now, the conversation has completely changed.
In one of the biggest opinion reversals seen in the AI industry, Musk has publicly admitted that he was wrong about Anthropic. Even more interestingly, the company is now one of SpaceX’s biggest infrastructure customers.
The shift highlights how quickly the AI landscape is evolving, where fierce competitors can also become valuable business partners.
Musk’s Public Change of View
In a recent post on X, Musk wrote:
“I was clearly wrong about Anthropic.”
He went even further by calling Anthropic the current leader in AI.
According to Musk:
- Anthropic currently has the strongest AI models available.
- Its latest Mythos and Fable models have set a new benchmark.
- He expects the company to release Mythos 2 soon, continuing its lead.
Coming from someone who has spent years criticizing the company, these comments attracted significant attention across the AI industry.
A Very Different Tone From Earlier This Year
The change becomes even more striking when compared with Musk’s previous comments.
Over the past year, he repeatedly criticized Anthropic.
Among his earlier remarks:
- Called the company “hypocritical.”
- Described Claude as “woke.”
- Accused Anthropic of promoting political bias.
- Claimed it disliked Western civilization.
- Suggested the company had little chance of winning the AI race.
- Even joked that Anthropic’s future was to become “misanthropic.”
At one point, Musk openly stated that winning was never a realistic outcome for Anthropic.
That makes his latest endorsement particularly significant.
What Changed?
One major reason is business.
Earlier this year, SpaceX and Anthropic announced a massive infrastructure partnership.
Under the agreement:
- Anthropic will rent computing capacity from SpaceX’s Colossus 1 AI data center.
- The contract runs through May 2029.
- Anthropic will pay approximately $1.25 billion every month for compute resources.
- The agreement provides access to over 300 megawatts of AI computing capacity.
- That includes more than 220,000 Nvidia GPUs dedicated to training and running advanced AI models.
The deal instantly became one of the largest AI infrastructure agreements announced so far.
It also shows how demand for computing power has become just as important as developing the AI models themselves.
Competitors Can Still Do Business
Some people questioned whether Musk could eventually stop supplying compute to Anthropic since they compete directly in AI.
Musk addressed that concern publicly.
Responding to a post suggesting that SpaceX could simply “cut off” Anthropic’s computing resources, Musk replied that although the companies compete, he would never intentionally hurt Anthropic by shutting down access.
He said that is simply not his style.
That statement reflects an important reality in today’s AI industry.
Companies increasingly compete in one area while partnering in another.
Just as cloud providers host competing software businesses, AI infrastructure providers can also support companies building rival models.
Musk Also Praised Anthropic’s Team
His change in tone extends beyond the company’s technology.
After spending time with Anthropic’s leadership team, Musk said he was impressed by both their competence and their commitment to building AI responsibly.
According to him:
- The leadership appeared genuinely focused on benefiting humanity.
- He found the team highly capable.
- He did not believe they were acting in bad faith.
For someone who previously questioned the company’s motives, this represents a significant shift.
The AI Competition Is Far From Over
Despite praising Anthropic, Musk has made it clear that the competition is only beginning.
He recently pointed out that:
- SpaceXAI is only about three years old.
- Anthropic has been building AI for roughly twice as long.
- OpenAI has an even longer head start.
Musk’s view is that today’s rankings may not determine who leads several years from now.
His message was simple:
Let’s see where everyone stands three years from now.
The AI Race Is Accelerating
The timing of Musk’s comments also comes during an unusually busy period for the AI industry.
Recent developments include:
- OpenAI unveiling GPT-5.6 with new flagship models for different user segments.
- Anthropic restoring access to its advanced Mythos and Fable models after regulatory restrictions were lifted.
- Anthropic launching Claude Sonnet 5, offering near flagship performance at a lower price.
- SpaceXAI preparing the public release of Grok 4.5, which Musk says delivers Opus-level performance with faster speed and lower operating costs.
The competition is no longer focused only on model intelligence.
Companies are also competing on:
- Cost
- Speed
- Infrastructure
- Developer adoption
- Enterprise customers
- Scalability
Winning in AI increasingly requires strength across the entire ecosystem.
Anthropic Is Also Expanding Beyond Chatbots
Anthropic recently introduced mobile and web versions of Claude Cowork, its AI agent platform.
The company also shared how customers are using the product.
The largest use cases include:
- Business operations and workflow automation
- Preparing reports
- Creating onboarding processes
- Managing spreadsheets
- Drafting presentations and proposals
Interestingly, Anthropic is increasingly positioning AI as a productivity assistant that handles routine work rather than replacing entire professions.
That messaging is softer than earlier warnings from CEO Dario Amodei, who had previously spoken more aggressively about AI-driven job disruption.
This broader shift mirrors comments from other technology leaders like Microsoft’s Satya Nadella and OpenAI’s Sam Altman, who have also adopted a more balanced tone around AI’s impact on employment.
What This Means for Investors
Musk’s reversal is important for several reasons.
First, it signals that Anthropic has earned recognition even from one of its strongest critics.
Second, it highlights the enormous value of AI infrastructure. Compute power is becoming just as strategic as the models themselves.
Third, it shows how interconnected the AI ecosystem has become. Companies that compete in AI development can still depend on each other for chips, data centers, cloud infrastructure, and distribution.
Finally, it reminds investors that opinions in technology can change quickly when products improve and business relationships evolve.
The Bottom Line
Elon Musk’s public admission that he was wrong about Anthropic is more than a change of opinion.
It reflects how rapidly the AI industry is evolving.
Just months ago, Anthropic was one of Musk’s favorite targets. Today, he calls it the industry’s leader, praises its team, and supplies the computing infrastructure powering many of its models.
Competition between OpenAI, Anthropic, SpaceXAI, Google, and other AI companies remains intense. But as AI becomes more capital intensive, partnerships may become just as important as rivalry.
In this industry, today’s competitor can also become tomorrow’s biggest customer.