Wire flash
Anthropic plans IPO around Nov 19, valuation seen up to $1.75 trillion
Editorial responsibility
- No named human review is recorded for this page.
- Source reporting is collected, normalized, translated or condensed automatically when needed.
- Automatically published source-backed update
Anthropic, the AI company known for its Claude language models and assistants, is preparing for a potentially record-breaking IPO, with first trading expected around November 19 and valuation estimates ranging from over $1 trillion to as much as $1.75 trillion. The company aims to raise roughly $64 billion, a figure only slightly below what SpaceX achieved in its summer listing. The article frames the IPO as a potential shift from the era of Big Tech to that of Big AI, raising concerns about market power and lock-in effects. Joris Kanowski of the Center for Digital Rights and Democracy argues Anthropic presents itself as different but essentially continues the Big Tech principle. The piece notes Anthropic's ties to Amazon, Nvidia, and Google, its large pre-purchased compute commitments, and a prior episode in which the US government ordered it to block its Fable 5 and Mythos 5 models for foreign users on national security grounds. It also cites political pressure from US midterm results and protests against data centers, with Kilian Vieth-Ditlmann of Algorithm Watch calling for clear regulation of data center construction.
Source report
What will this cost! Construction of a data center in Nevada. Photo: Daniel Cole/reuters
By Svenja Bergt
The document that could herald a new era does not begin with words, but with three images. Each shows half of a petri dish, a fossil, and an astronomical map. For anyone who doesn't immediately grasp the message, Anthropic explains directly below what this rather unusual introduction to a stock prospectus is meant to symbolize: "We believe that AI is one of the most important technologies in human history." Bam! On the other hand, what else should an AI company on the verge of a historically large IPO write?
Over 1 trillion US dollars, up to 1.4 trillion, perhaps even 1.75 trillion—the estimates for the valuation Anthropic can expect as a starting figure open up new dimensions. The first day of trading is expected to be November 19. With the shares sold, Anthropic—best known for its language models and AI assistants under the name Claude—aims to raise around 64 billion US dollars. This puts the forecasts only slightly below what the space company SpaceX achieved in its IPO this summer.
But this is not just about a new financial dimension. It is also about a new dimension of market power. A sense of what consequences this can have could be observed this summer. At that time, the US government instructed Anthropic to block its powerful models "Fable 5" and "Mythos 5" for foreign users and employees—for reasons of national security. For Anthropic, which sees itself as a good AI company committed to the common good and wants to be seen as such, implementing this instruction would have seriously damaged its image. It therefore argued that it could not distinguish between domestic and foreign users and simply blocked access for everyone.
wochentaz
This text first appeared in wochentaz, our weekly newspaper from the left!
Every week, wochentaz is about the world as it is—and as it could be. A left-wing weekly newspaper with a voice, a stance, and the distinctive taz view of the world. New at newsstands every Saturday and of course by subscription.
The episode is almost half a year old. The excitement about what consequences it would have had if the models had been on the market longer and companies and governments had relied on being able to access them has somewhat subsided. But Anthropic's upcoming IPO raises the question again. Is a Big AI now emerging after Big Tech—with all the negative consequences that such power brings?
"Anthropic always acts as if they were different from the others. That may be partly true, but essentially it is the continuation of the Big Tech principle," says Joris Kanowski, a consultant at the Center for Digital Rights and Democracy. By "different from the others," he refers to Anthropic's history. The company was founded in 2021 by former OpenAI employees. They wanted to present an entrepreneurial counter-model. "Our guiding principle is to develop and maintain advanced AI responsibly, for the long-term benefit of humanity," states the current public benefit report.
Sparks of this idea show up, for example, when the company resists the US government using Anthropic's systems for mass surveillance and in autonomous weapons systems. Nevertheless, Anthropic is intertwined with major tech corporations such as Amazon, Nvidia, and Google. And it is under pressure. According to its stock prospectus, it has already purchased computing power and infrastructure worth several hundred billion US dollars for the coming years. The money first has to come in.
A System Without Escape
"I believe Anthropic is trying to do something similar to Google, from whose ecosystem you can no longer escape," says Kanowski. "And with AI it will be even more extreme." Because with AI, dependencies can become even greater. If people use AI chatbots in everyday life to understand official letters and do their tax returns, if companies program their software with AI assistants, if AI is in all smartphones and networked devices, and if doctors use it for documentation and diagnostics—then there is quickly a problem if that suddenly disappears. For example, because the corresponding AI model is blocked for foreign countries under pressure from the US government.
The good news: This development is by no means inevitable. While the search engine market is maximally concentrated and Google holds around 90 percent, the AI models market is still comparatively open. On the one hand, there are the major US providers like Anthropic and OpenAI, which dominate extremely powerful models. On the other hand, there are also Chinese companies like Deepseek and Moonshot AI pursuing an alternative approach. Their models are partially public and can therefore be operated and used independently of the original provider, which enables price competition.
And then there are European providers like the French Le Chat from Mistral or Apertus from Switzerland. The latter is an open-source model that institutions and private individuals can not only operate themselves, but whose training data is also openly available.
In the days after the midterms in the US, the stock prices of many tech corporations fell.
"We must ensure that the AI models market remains open," says Joris Kanowski. This will not happen automatically. After all, the US providers have a great interest in concentration. The less competition, the sooner they reach profitability. And the sooner there are lock-in effects. This means that users who have once committed to a particular provider find it difficult to get out. A problem that authorities and companies working with Microsoft products are familiar with today.
What could now slow down growth—currently designed for maximum—is precisely the current political changes in the US. In the midterm elections, the Democrats won majorities in numerous previously Republican-governed states. And they view the excesses of tech growth more critically than the Republicans. Consequently, the stock prices of many tech corporations fell in the days after the election, but have since recovered.
Movement Against Data Centers
Political pressure to provide more regulation comes from a protest movement against data centers that has formed in many states. "Policymakers cannot close their eyes to this pressure," says Kilian Vieth-Ditlmann from the NGO Algorithm Watch. A first step of regulation would be clear rules for the construction of data centers—for example regarding location, power supply, and sustainability. He assumes that this would also contain the market dominance of tech corporations in AI. Because even though Vieth-Ditlmann, like Joris Kanowski, considers the models market still reasonably open, this is no longer the case when it comes to the economic significance of the companies.
According to analyses, tech investments—especially the construction of data centers—now significantly drive US economic growth. This dominance can become a risk if a bubble forms here. And there are some signs of that: the close interconnection of many corporations, for example. Nvidia, the market-leading manufacturer of high-performance chips, is not only a supplier to Anthropic but also an investor and customer of AI services. If one player in such a structure falters, it can affect the entire industry. And that is happening right now. On Thursday, Nvidia CEO Jensen Huang, who is accused of an affair with an alleged spy from China, resigned from his position.
Also against the backdrop of a possible bubble, Kilian Vieth-Ditlmann views German plans for rapid data center construction critically: "There is a strong FOMO in Europe right now." A fear of missing the connection, of not being part of it when the AI boom elsewhere provides economic growth, jobs, and prosperity.
If. Because ultimately, according to Vieth-Ditlmann, being fast also carries risks; for example, if overcapacity in data centers arises that cannot be refinanced: "We don't need to replicate the mistakes of the US."
A financially closely intertwined industry, mutual dependencies, bubble risk, immense investments, cyberattacks by AI agents, political uncertainties, and now also an alleged espionage affair—quite a lot of risks. Anthropic also lists the dangers of the technology in general and the risks for the company specifically over almost a hundred pages in its stock prospectus. After the slight stock market turbulence following the midterms, there was already speculation about whether Anthropic would postpone the IPO. The argument against it: market conditions might not get any better. Perhaps the IPO also heralds the following upheaval: the beginning of the end of the great AI hype.
Related articles:
- Espionage Affair at Nvidia: What Jensen Huang's Downfall Means for China — In the People's Republic, the alleged espionage affair of Nvidia CEO Huang is being celebrated. In the long term, however, the scandal is likely to harm Beijing. By Fabian Kretschmer
- Portrait of Jensen Huang: The Master of AI and His Male Drive — Nvidia CEO Jensen Huang admits to an affair with a Chinese woman whom the US intelligence service considers a spy, about whom he warned as early as 2025. By Sven Hansen
- What Makes Nvidia So Important: The Hardware Behind AI — AI chatbots in their current form would be inconceivable without Nvidia's chips. Germany also wants to be close to the US corporation. By Svenja Bergt
Source
taz.de - taz.deWestern
Part of this Story
Anthropic sets $70-$78 IPO range, targeting up to $1.4 trillion Nasdaq valuation