Anthropic, the creator of the Claude language model, has become the most valuable AI startup in the world. After its latest funding round, in which it raised $65 billion, it reached a valuation of $965 billion — and for the first time in history, it surpassed its main rival OpenAI, which was valued at $852 billion in March. Anthropic has been operating for only half the time it took OpenAI to reach its current status. Alongside the funding announcement, the company launched the Claude Opus 4.8 model and confirmed the upcoming release of the powerful Mythos for cybersecurity.
Listen to this article:
How Anthropic overtook OpenAI: numbers that rewrite the rankings
On May 28, 2026, San Francisco-based startup Anthropic announced it had raised $65 billion (approximately CZK 1.43 trillion) in its latest funding round. The company's new valuation reached $965 billion, allowing the creator of Claude to surpass OpenAI, whose value based on its last round in March 2026 stood at $852 billion. Both companies could go public as early as 2026, which would represent one of the largest waves of tech IPOs in history. The pace at which Anthropic grew is remarkable. OpenAI, which kicked off the generative AI boom in 2022 with its ChatGPT chatbot, needed roughly a decade to reach its current valuation. Anthropic — founded only in 2021 — achieved an even higher value in half the time, The New York Times noted. It took just five years.Who is behind the record funding
The latest investment round was led by four major venture capital funds from Silicon Valley: Altimeter Capital, Dragoneer, Greenoaks, and Sequoia Capital. The round also includes $15 billion in previously committed investments from cloud giants, including $5 billion from Amazon, which is a long-term strategic partner of Anthropic. Among the investors, key semiconductor manufacturers appeared for the first time — Micron, Samsung, and SK Hynix — whose chips are absolutely essential for training and running large language models. The strategic entry of hardware manufacturers into Anthropic signals that investors view AI infrastructure as a long-term promising business, not a short-term speculative bubble.Why Anthropic succeeded: a bet on enterprises and AI safety
While OpenAI initially bet primarily on the mass adoption of ChatGPT among everyday users, Anthropic targeted enterprise clients from the start. This strategy has proven exceptionally lucrative. Corporate customers pay for reliability, safety, and consistent performance — and these are precisely the values on which Anthropic has built its brand. The company was founded in 2021 by former OpenAI employees, including current CEO Dario Amodei, who previously led the research department at OpenAI. The team left the parent company partly due to concerns about insufficient emphasis on AI safety. From day one, therefore, Anthropic has placed the topic of responsible AI development at the center of its corporate identity — and as the numbers show, the market values this bet. CFO Krishna Rao said of the latest round: "This funding will help us serve the historic demand we're facing, stay at the forefront of research, and get Claude into more places where real work gets done." In other words: Anthropic is betting that companies want to integrate AI directly into their workflows, not just chat with it.Claude on all three major clouds
Anthropic also announced that Claude has become the first frontier AI model available on all three largest cloud platforms: Amazon Web Services (AWS), Google Cloud, and Microsoft Azure. This broad availability gives Anthropic a significant competitive advantage — companies can run Claude on the infrastructure they already use, without needing to change vendors or migrate data.Claude Opus 4.8: honesty as a new feature
Alongside the announcement of record funding, Anthropic launched a new version of its most powerful model — Claude Opus 4.8. The main novelty this time is not higher performance or speed, but improvements in honesty. The model more often signals when it is unsure about something and less frequently generates unfounded claims. "A general problem with AI models is that they sometimes jump to conclusions prematurely and confidently claim they have made progress, even when they lack sufficient evidence," Anthropic explained. The new Opus 4.8 addresses this weakness — and according to early testers, it is noticeably more cautious in situations where it would previously "hallucinate." The model is available at the same price as its predecessor and shows slight but measurable improvement across benchmarks. For enterprise deployment, this shift is crucial: unfounded AI claims can cause real financial or reputational damage in an enterprise environment.Mythos and Project Glasswing: AI as a cyber threat hunter
Anthropic confirmed that it plans to make its strongest model Mythos available to all customers in the coming weeks. Mythos is a large language model with advanced cybersecurity capabilities that, during testing, managed to detect over 10,000 critical software vulnerabilities — including bugs that human developers had overlooked for decades. As part of the Project Glasswing initiative, technology companies such as Amazon, Microsoft, and Apple have already gained access to Mythos, using it specifically to bolster cyber defenses. However, Mythos's capabilities also raise concerns among many experts and policymakers — advanced cyber tools in the hands of an adversary could pose a fundamental security threat.A fierce clash with the Pentagon
Interesting context for Anthropic's growth comes from the ongoing legal dispute with the U.S. Department of Defense. The Pentagon labeled the company a "supply chain risk" — a move Anthropic challenged in court as unconstitutional retaliation for refusing to grant the military unrestricted access to its AI models. Anthropic's stance stands in direct contrast to the approach of OpenAI, Google, or Nvidia, which actively collaborate with the Pentagon on classified projects. For a company that builds its brand on responsible AI, this position is consistent — but also risky at a time when military contracts make up a significant portion of the tech sector's revenue.SpaceX xAI: Elon Musk targets $1.75 trillion
Anthropic and OpenAI are not the only players in the race for the title of most valuable AI company. SpaceX, which absorbed Musk's AI startup xAI in February 2026, could go public as early as June 12, 2026. The target valuation? $1.75 trillion — which would make it the largest initial public offering (IPO) in history, far exceeding even Anthropic's current value. If all three planned IPOs take place, Wall Street will experience an unprecedented wave of AI capital in 2026.What it means for the Czech Republic and everyday users
Claude is readily available to Czech users as well — via the web interface at claude.ai or the mobile app. The model supports Czech, although English remains the primary language, and in the Czech environment you may occasionally encounter minor nuances in understanding local specifics. For everyday communication, translations, or text analysis, however, it is fully usable. For Czech companies, it is key that Claude runs on all three major cloud platforms that are standardly available in the country (AWS, Google Cloud, Microsoft Azure). This means companies can integrate the model into their existing workflows without needing to change infrastructure. The broader impact is strategic: growing competition among OpenAI, Anthropic, and other players creates pressure on both quality and price. For European businesses, this means more options when choosing an AI provider and a stronger negotiating position. At the same time, it underscores how far Europe is from a leadership position — not one of the top AI startups was founded in the EU. With upcoming regulations under the EU AI Act, it will be interesting to see how American companies adapt to European requirements for AI transparency and safety.Will Anthropic become a publicly traded company and how will it affect Claude users?
Yes, according to available information, both Anthropic and OpenAI could go public during 2026. For everyday Claude users, an IPO should not mean immediate change — the free tier remains. In the long term, however, going public usually means greater pressure on profitability, which could be reflected in subscription prices or the scope of features in the free version. Specific plans have not yet been disclosed.
How does Claude's Czech language support compare to ChatGPT?
Claude supports Czech and can communicate in it at a very good level. In a direct comparison with ChatGPT, the quality of Czech is comparable, although ChatGPT, thanks to its longer history and larger user base, has slightly better understanding of Czech-specific contexts and particular linguistic nuances. For everyday use in Czech — whether writing, translation, or analysis — both models are fully sufficient.
Why did Anthropic refuse to provide its models to the Pentagon and what does this mean for AI safety?
Since its founding, Anthropic has emphasized a responsible approach to AI development and an effort to minimize risks of misuse. Refusing unrestricted access to the Pentagon is consistent with this philosophy — the company does not want its models to be used for military or intelligence operations without sufficient oversight. This stance clearly differentiates Anthropic from the competition, but also puts it in a difficult position, as the Pentagon reacts by labeling the company a supply chain security risk.
