This is insane. Anthropic is on fire. Its global market share has grown by 70 percent. It has taken 30 percent of OpenAI’s user base. The craziest part is that its annual revenue has reached 20 billion dollars. It doubled in just two weeks and added 5 billion dollars. Two and a half million people voted with their wallets and chose Claude.
But Anthropic did not just win. It won in a very ugly way.
“The whole world has been refreshed by one picture.”
In February 2026, Anthropic’s global market share grew by 70 percent and quickly surpassed OpenAI.
In just a short time, ChatGPT’s original 90 percent market share was taken over by Claude.
What is even more shocking is that Anthropic’s annual recurring revenue has once again broken records. It is approaching 20 billion dollars.
In just two weeks, it added 5 billion dollars.
According to the Fortune Top 500 list, it ranked 166.4. If you look at this data, you can see that Anthropic’s growth is really too fast.
A Stunning Reversal: 20 Billion Dollars
Anthropic’s revenue growth is truly jaw-dropping. After all, just a few months ago, it was still a “small company” that people were worried about.
Some people even said that Anthropic was “a company that is half water and half fire.”
On one hand, Anthropic’s ability to make money is skyrocketing.
CEO Dario Amodei revealed that the company’s annual recurring revenue has reached 20 billion dollars.
It added 5 billion dollars in just two weeks. If you look at this growth rate, it is enough to make any investor’s eyes light up.
On the other hand, the market is still very worried about Claude’s ability to keep growing.
After all, in 2026 alone, Claude Code exploded, the flagship Opus 4.6 was released, and Claude grew hands and feet.
In February, Claude Cowork became a killer feature. In just ten days, it completed a major update. It brought a huge boost in value to the whole world.
According to more detailed data, Anthropic is currently valued at 300 billion dollars. Its valuation has directly jumped to 380 billion dollars.
At the same time, its annual revenue run rate has reached 14 billion dollars. This means it is growing by more than 10 billion dollars every month. It is an explosive growth rate.
Some statistics show that Anthropic’s business is very healthy.
In the past year, customers who spend more than 10 million dollars on Claude have grown by 7 times. Customers who spend more than 100 million dollars have grown by 10 times. The top ten customers have grown by 500 percent.
Among the Fortune 10 companies, 8 of them have become actual users of Claude.
Since Claude Code was released in May, this agent tool has directly generated 25 billion dollars in revenue.
According to user feedback, Claude Code is now being used by the whole developer community. Its usage rate is growing every day.
Here is a shocking piece of data. Four percent of all GitHub code commits are now automatically generated by Claude Code. That is one out of every 25 commits.
Claude Code has become the AI tool that the world’s largest companies rely on to write code.
Top companies like Shopify, NASA, Figma, and Stripe have all paid money to Anthropic.
A Total Victory
According to Deedy’s data, Anthropic has achieved a “total victory.”
Similarweb statistics show that Claude.ai’s monthly active users have been growing rapidly. It has surpassed Grok and DeepSeek.
The latest Ramp AI index shows that Anthropic’s business share has grown from 16.7 percent to 19.5 percent. OpenAI has dropped to 35.9 percent.
Among them, one out of every five companies has chosen Anthropic. The ratio was 25 to 1 before.
In the API market, Anthropic now holds 90 percent of the market share. It is far ahead.
While Anthropic is building a skyscraper, it also faces a huge challenge.
On the Google Play store, the Claude app has been removed from the top downloads list. On the App Store, it has also been taken down.
Some people say that users are leaving. Others say that Claude is being completely blocked.
However, behind this ban, Anthropic has been pushed to the “edge.”
“Because of AI safety alignment issues, the government has directly banned Claude. The web version and the app have all been blocked.”
Defense Secretary Pete Hegseth has signed an order to ban the use of Claude in the military. This has had a huge impact on Anthropic’s business.
At the same time, it has also put pressure on other AI companies.
The Pentagon Ban: The First Major Product Ban
According to a report by The Information, the partnership between Anthropic and Palantir may become the “first major product ban” in this war.
For a long time, Anthropic has been working with defense giant Palantir to provide services for the military.
In Palantir’s system, Claude models are deployed.
The Pentagon’s ban directly affects Claude’s military use cases. This is a key blow.
However, the ban is not just about national security. It is also about “data leakage” risks.
The new regulations clearly state that the military cannot use any Anthropic technology.
Last year, 42 percent of Palantir’s revenue came from government contracts. For Palantir, this is a “double-edged sword.”
Although Anthropic’s revenue is growing very fast, losing Palantir’s government business is a very serious blow to its market position.
The Pentagon has always been Palantir’s customer, but it uses Claude managed by AWS.
It is worth noting that Palantir’s internal systems also use “fetal” models.
When the ban was first announced, Microsoft was considering replacing Claude with OpenAI or other models. This also affected the partnership.
On one hand, Claude users around the world are experiencing a 6-month “painful transition period.”
According to people familiar with the matter, Palantir CEO Alex Karp has been communicating with the Pentagon about this issue.
He revealed that the Pentagon has given two options. One is to continue using the current model. The other is to switch to a new model. But switching will only lead to more problems.
Karp strongly supports the first option. He believes that switching will cause more trouble.
Palantir currently allows clients to choose different AI models on its defense platform. They can choose between Anthropic, OpenAI, and other providers.
However, an Axios article pointed out that the Pentagon’s ban on Anthropic is actually a “joint ban” by the British and American governments.
In theory, both countries’ military and intelligence agencies can use it. But free ai hentai generator in practice, the product supply chain and the review system are different.
In this war, OpenAI quickly positioned itself as a “replacement.” It joined hands with GPT. But in the end, it also lost something.
OpenAI’s Self-Restriction: The Real Loser
But today, an internal document from OpenAI has once again revealed some details about its partnership with the government.
The hentai ai chat document shows that OpenAI’s partnership with the government is not as simple as it seems. It involves a series of complex rules and restrictions.
The document specifically emphasizes that OpenAI must obtain government approval at multiple levels before it can access private data. It must also use official AI tools in government work.
sextbots
What is interesting is that these restrictions also apply to safety. The NSA report shows that government agencies are currently prohibited from using OpenAI’s tools.
These agencies need to sign a revised contract with GPT before they can use it.
In this war, OpenAI is actually the biggest loser.
Because the high-level restrictions are too strict, government agencies cannot use it. This directly affects the entire market.
In fact, OpenAI is also facing a similar problem internally.
Many employees are worried about the company’s safety. They have never seen such a strict attitude. Some employees even said that OpenAI is “self-restricting.”
Some employees directly said that OpenAI cannot meet the government’s security requirements. It cannot use its own tools.
Defense Secretary Pete Hegseth also said the same thing.
What is even more interesting is that OpenAI’s “original sin” is that it has been trying to restrict its own employees. But in the end, it has restricted itself.
2.5 Million People: The Movement to Quit ChatGPT
But the real problem is not just about ChatGPT being banned. It is about the users leaving.
The global “Quit ChatGPT” movement has officially started. A total of 2.5 million people have joined.
On Reddit, quitting ChatGPT has become a new trend. It is the “correct” choice.
A while ago, OpenAI completed an 11 billion dollar funding round. But according to data from QuitGPT, the number of weekly active users has not increased much.
On Instagram, a post about this topic suddenly got 36 million views. It was shared 1.7 million times. The whole internet is talking about it.
NYU marketing professor Scott Galloway directly said that the only way to vote is with money. He said that these tech giants have too much power. They are causing pain to ordinary people.
In 190 countries, 2.5 million people have joined the “movement to quit AI.” The future of the AI industry has become even more difficult.
Anthropic’s “total victory” is not just a business victory. It is also a victory for OpenAI’s “self-restriction.”
This war of power, money, and original sin is like a historical poem. It is being written in real time.
Who will win in the end? No one knows.
CADOAN is a professional, independent AI industry blog and information platform dedicated to the research, sharing, and popularization of artificial intelligence. We are a team of AI enthusiasts, researchers, and technical writers who focus on the development and application of modern artificial intelligence. We do not represent any commercial institution, technology company, or AI model camp. Our only position is to provide real, objective, and valuable AI content for readers, learners, developers, and business practitioners around the world.