OpenAI's unexpected rival. How a tech company got to a $1 trillion valuation when it seemed like it had no chance

Two years ago few had heard of the name Anthropic and it seemed that OpenAI's position with its ChatGPT would hardly be contested. Well, in recent months the Anthropic company has grown amazingly in value and notoriety and ended up being valued at $1 trillion. Along the way he's feuded with the Pentagon, hired a female philosopher, and scared the world with an AI model too powerful to roll out to everyone.

In 15 months, Anthropic's company value increased 15 times, but the beginning was complicated, Anthropic was founded by seven former employees of OpenAI. Between the two companies, the rivalry intensified, especially since in the early years it seemed that Anthropic had no chance in the fight with the giant led by Sam Altman and made famous by ChatGPT, the first chatbot that made a sensation among the public.

The word “anthropic” refers to something about man or mankind. For example, in geography, they learn about “anthropogenic erosion” in school.

An “Ethical AI” – A different tactic from the competition

Anthropic was created in San Francisco in 2021 by brothers Dario and Daniela Amodei and was initially known for the chatbot Claude, launched in March 2023 in the US and in 2024 in Europe. In recent months she has become famous for the dispute with the Pentagon and the model Claude Mythos.

The Claude chatbot has two types of monthly subscription: one that costs 18 euros and a more advanced one of 90 euros. The language model promises nuanced conversation management and excels at searching for information in context and provides a high degree of personalized support.

Anthropic is trying to distinguish itself from its competitors by implementing stricter safety measures and ethical principles for this AI model technology that is causing both excitement and concern.

That we are not talking about an ordinary company also depends on the fact that Anthropic has among its employees a woman who is a philosopher and who works on the “character” of the chatbot Claude. Amanda Askell is her name and she builds the Claude “personality” to decide how the model responds (in terms of tone, style, empathy and level of caution). She also works on how the AI ​​behaves in sensitive situations, helping to establish a set of rules. Here you can see an interview with Amanda Askell.

Critics of the company say there is also a lot of marketing in this image of “ethical AI”. Some note that we must not forget that Anthropic is still a commercial company that builds increasingly powerful models and earns ever-increasing revenue.

One step away from the first quarter with profit

The latest data indicates that Anthropic will turn an operating profit for the first time in the April-June quarter, while OpenAI doesn't expect to turn a profit until 2030. Moreover, revenue is expected to hit $10.9 billion in the current quarter, up from $4.8 billion in the first three months of the year. The press wrote that compared to OpenAI, Anthropic was able to train AI models at lower costs than its rivals.

How did the company get this far? Anthropic launched competitive AI models, focused on enterprise customers, and – because of its strong emphasis on the safety and reliability of AI systems – this attracted the trust of investors and users.

The company has also come a long way thanks to the skill of CEO Dario Amodei, who left OpenAI after falling out with the company's management and the way AI technologies developed by OpenAI were funded and launched.

The personal rivalry between Dario Amodei and Sam Altman has moved into the public and geopolitical space. At a major summit in India in February, the two CEOs pointedly refused to shake hands during the official group photo, preferring instead to awkwardly touch elbows.

Additionally, Anthropic launched an aggressive advertising campaign during the Super Bowl finale, directly attacking OpenAI's plans to introduce ads into ChatGPT with tongue-in-cheek spots and the final tagline: “Ads are coming to AI. But nt to Claude.”

Dario Amodei said in a 2023 interview that there was a 10% to 25% chance that AI technologies would destroy humanity, but in October 2024 he was much more optimistic and published a 14,000 word essay on the benefits of AI. The article was titled: Machines of Loving Grace – How AI Could Transform the World for the Better

Anthropic also has a lot of marketing in its overall strategy and knows how to communicate in a way that gets the attention of the whole world. The company warned that the technology it was developing could be dangerous to humanity, something few companies would dare to say publicly. The statements not only sparked controversy, but – on their occasion – a lot of people heard about the company and some opened Claude for the first time.

Just the other day Jack Clark, the co-founder of the company Anthropic, said in an interview with the BBC that there must also be the possibility of slowing down the evolution of artificial intelligence (AI), warning that the technology is approaching a point where it could develop without human intervention.

“You want to have the option to take your foot off the accelerator and put your foot on the brake,” Clark told BBC Newsnight. “Right now, it's like the AI ​​industry has an accelerator but no brake.”

Masters of image shots

The focus on safety, however, helped the company maintain a positive public image, but it also created tensions. This year, the Pentagon banned the use of Anthropic technology in its own systems because the company wanted to impose some limits on how military and intelligence officials could use the “made by Anthropic” technology.

The White House has labeled the company a national security risk, and President Trump has ordered federal agencies to cut ties with Anthropic. Since then relations with the US authorities have improved a lot.

From the whole story with the Pentagon, Anthropic came out with an image gain because public opinion turned in favor of the company. Its leaders were perceived as willing to stand up for their principles and not compromise at all, even at the cost of losing huge contracts. This contrasted with the position of OpenAI, which had a reputation as a company that many said had “sold its soul” for handsome financial gains.

In April, Anthropic took another image hit after announcing the development of a super-sophisticated model called the Claude Mythos.

Anthropic said this model is so strong in cybersecurity that it decided not to release it publicly, and only a few entities were able to access it as part of Project Glasswing. The Mythos has sparked backlash from central banks and intelligence services around the world, with the main upset being that Anthropic is the only one who decides who has access to this extremely powerful model.

Eventually the company said it would expand the number of entities that will have access to Mythos and recently promised that it will also be available in the European Union. If at first only 40 entities had access to Mythos, the company recently announced that the number will reach 150 organizations.

Anthropic said this model has an almost unbelievable ability to find and exploit hidden vulnerabilities in the software that controls banks, power grids and other critical infrastructure. It has become a geopolitical stake full of controversy, especially since an American company is in control.

A “dance” of billions of dollars

Anthropic reached the $1 trillion valuation also thanks to major investments from tech giants, and its latest funding round in May 2025 was $65 billion, while the March 2025 round was “only” $3.5 billion.

Amazon, Google, Microsoft and Nvidia are among the big tech companies that have invested billions in rival OpenAI. Large investment funds also have a strong presence, whether we are talking about American ones, some with an international presence or investors from the Arab world.

The company grew its business and because it kept expanding: India became the second market for the use of Claude, and in recent months Anthropic announced its expansion in Europe as well, the most recent office being in Milan, the Company had five European offices: London, Dublin, Zurich, Munich and Paris.

How much Anthropic has grown in value in 15 months

• March 2025: $61 billion
• September 2025: $183 billion
• February 2026: $380 billion
• May 2026: $965 billion

How Anthropic came to be worth $1 trillion

It is a “textbook” example of how to build a successful company in five years, but the “recipe” cannot be applied by any company, because the ingredients were “premium” and the time chosen to start the road was right.

Anthropic started with a team of exceptional specialists who understood that these large language models were going to grow a lot and also understood that they had to do something different from the #1 in the market.

The company has focused on programming and integrating its apps into the enterprise environment, and it's gotten to the point where a good portion of its revenue comes from enterprise customers, not ordinary people who pay a subscription to chat with the chatbot. Claude AI is among the best chatbots for code writing and analysis.

Last but not least, the company has focused on the ethical use of AI, and there is also a “Claude Constitution” that talks about the “character” of the chatbot. Company bosses argue that it is possible to build highly capable systems that are – simultaneously – useful, honest and secure.

Anthropic claims that it tries to train the model using this “constitution”, meaning an explicit set of principles and values. The model learns to critique and correct its own responses based on these principles, the company says.

Critics note that it is unclear who decides what values ​​go into the “constitution,” and it remains an open debate whether a private company should set such principles on its own that have broad societal implications.