Anthropic files confidential IPO application, outpacing OpenAI in the race to go public

Race with OpenAI

With the confidential filing, Anthropic is gaining a head start over rival OpenAI, according to observers. Insiders report that OpenAI is also preparing a confidential application in the coming weeks. The race between the two leading AI developers for the industry's first major IPO is thus considered open.

Anthropic was founded in San Francisco in January 2021. The founders are siblings Dario and Daniela Amodei, who previously worked at OpenAI and left the company due to differences in opinion with then-OpenAI CEO Sam Altman. A central point of contention was the issue of AI safety. The company's headquarters are located in an office building on Market Street in San Francisco.

Valuation Surpasses OpenAI

Anthropic's valuation has risen rapidly in recent months. In February 2025, it was still at $380 billion; in May 2026, after a funding round of $65 billion, it reached a valuation of $965 billion. This represents a fifteenfold increase within 14 months. According to estimates, Anthropic has thus also surpassed the valuation of OpenAI, which was most recently reported at $852 billion.

Anthropic pursues a different business strategy than its main competitor. While OpenAI primarily targets private end-users with its chatbot ChatGPT, Anthropic focuses on paying business customers. The company generates its largest revenue from AI software that automatically generates computer code. The tool "Claude Code" enables users to create software without programming knowledge.

Business Model: Focus on Enterprises

The gap is also evident in revenue. Anthropic's annualized revenue is currently estimated at $45 billion, while OpenAI's is around $30 billion. In the full year 2024, Anthropic still generated $10 billion in revenue; for 2026, forecasts based on current developments are trading at $47 billion. The company expects to become profitable for the first time in the second quarter of 2026, having previously expected this no earlier than 2028.

Anthropic's investor base is broadly diversified. Early backers included former Google CEO Eric Schmidt, Skype co-founder Jaan Tallinn, and IT entrepreneur Dustin Moskovitz. Since then, the company has completed around eight funding rounds. Amazon and Google each hold an estimated 15 percent of the shares and simultaneously provide the company with computing power. Microsoft is also among the compute suppliers, albeit to a lesser extent.

Computing Power as a Bottleneck

An indication of the demand for skilled labor: Anthropic employs an estimated 3,000 to 5,000 employees. Very recently, Andrej Karpathy, co-founder of OpenAI, moved to Anthropic. The company also operates its own consulting firm that trains third-party companies in handling AI software.

Anthropic operates a new AI model called "Mythos," which is designed to detect security vulnerabilities in computer systems. According to the company, the cost of using Mythos is six times higher than its previously most advanced model. The system is not freely accessible but is only available to a select group of users. Both US authorities and the EU have expressed interest in test access to evaluate potential risks.

Anthropic is also taking unusual steps regarding computing power. Due to enormous demand, the company has had to temporarily throttle quotas for customers. A few weeks ago, Anthropic signed an agreement with SpaceX and its AI subsidiary xAI to utilize unused capacity from associated data centers. These data centers have faced criticism in media reports for using gas turbines for power generation without sufficient consideration for local residents.

Risks and Valuation Concerns

Despite strong business development, not all observers are optimistic about the IPO. Some experts express doubts about whether the hundreds of billions of dollars currently being invested in AI infrastructure can be recouped in the long term. Furthermore, there is currently no real lock-in for companies with individual AI providers: firms can switch to another provider at short notice if it delivers better results.

Observers are divided on the strategic significance of Anthropic's move. One group of analysts sees the company at an advantage because its earlier filing gives it the power to define which company has the better market position. The other group argues that OpenAI benefits because it can wait to see how institutional investors react to the figures of a leading AI company.

According to reports, Anthropic's IPO could take place in the fall of this year, but no earlier than after the SEC review is completed. Until then, both the size of the share offering and the specific terms remain open. The reporting is based, among other things, on research by Nils Dampz from the ARD studio in Los Angeles.

Outlook for the Fall

Regardless of the outcome of the process, the filing is considered a signal of the AI industry's maturing process. For the first time, investors could directly participate in the economic success of a pure AI developer that is in direct competition with the hitherto dominant provider OpenAI. The coming months will show whether the high valuation level in the capital market will hold and whether the industry's investment cycles will indeed bear fruit.