When Anthropic goes public this month, in what is expected to be one of the largest IPOs (initial public offerings) in history, the company will create new billionaires and inject a wave of wealth into its headquarters city, San Francisco, in the state of California (USA).
But some of the biggest beneficiaries will be tech giants already worth trillions of dollars. Amazon and Google invested in Anthropic and are its largest external shareholders, with over 25% of the company's shares together.
If the IPO reaches the value expected by Anthropic, the company could be valued at up to US$1.4 trillion (R$7.7 trillion). Amazon's investment of US$18 billion (R$99 billion) could become worth more than US$250 billion (R$1.375 trillion), while the US$33 billion (R$181.5 billion) invested by Google could reach US$140 billion (R$770 billion).
The size and speed of these gains are unprecedented in the history of corporate investment in technology companies. They also reflect a paradox of the artificial intelligence boom: although startups are at the forefront of AI development, the sector's biggest companies are the ones profiting most from demand for the technology.
Gil Luria, technology analyst at investment bank D.A. Davidson, estimates that more than US$1 trillion (R$5.5 trillion) in market value for each of the companies — largely the result of offering cloud computing to the startup and marketing its AI to customers — arises directly from the relationship with Anthropic.
Artificial intelligence is "an extraordinary opportunity to forever change the size" of Amazon, the company's chief executive, Andy Jassy, told investors this year.
Tech giants were in a position to benefit from Anthropic's rise because they operated some of the most profitable businesses in the world. Like its main competitor, OpenAI, Anthropic, founded just five years ago, needed huge capital infusions to finance the development of its AI. Only a small group of companies were able to provide so much money.
Even Nvidia, today one of the most valuable companies in the world, didn't have so many resources a few years ago. The company simply wasn't "in a position to make a multi-billion dollar investment in Anthropic so that it could use our computing power," Nvidia Chief Executive Jensen Huang said in a podcast this year.
The bets on Anthropic occurred when two factors met. After its founding in 2021, the company's first investors were mainly individuals with large fortunes, including Jaan Tallinn, co-founder of Skype, and Sam Bankman-Fried, an entrepreneur in the cryptocurrency sector. After the collapse of Bankman-Fried's company, amid a scandal and bankruptcy proceedings, Anthropic's future was uncertain.
A few weeks later, OpenAI quietly launched ChatGPT and began the race for artificial intelligence.
Amazon was an early backer of OpenAI when the company started as a nonprofit in 2015, but dropped out a few years later when the startup was looking for a large partner to finance its plans. Instead, Microsoft invested US$1 billion (R$5.5 billion) in OpenAI in 2019.
The initial success of ChatGPT and Microsoft's appreciation associated with the tool's advancement triggered a race among other technology companies.
Large companies saw strategic advantages in partnering with Anthropic. They provided advanced chips and computing capacity to the startup and gained access to the company's systems to offer them to their own customers.
Google made its first investment in Anthropic in 2023, and the startup committed to using the company's internally developed chips. In September of that year, Amazon, which was facing the perception of being behind in the AI race, made its first contribution, US$1.25 billion (R$6.87 billion). Around the same time, Anthropic chose Amazon as its primary provider of cloud computing services.
The relationship was "very symbiotic," Adam Selipsky, who ran Amazon's cloud services division, said in an interview that fall. "The people at Anthropic are very smart — some of the best technical minds in the world."
Anthropic gave Amazon a cutting-edge customer, capable of pushing the giant to radically improve its internally developed AI chips, which seek to compete with those from Nvidia. In return, Amazon committed to providing the startup with enormous volumes of computing capacity.
Selipsky called the equity investment in Anthropic a "natural offshoot of this very, very strong business relationship."
For Jassy, the investment represented a low-risk bet because Amazon could advance its cloud services partnership with Anthropic, said Matt McIlwain, managing partner at Seattle-based Madrona Venture Group.
"The potential gain," McIlwain said, "came in the form of a significant equity stake."
Marc Benioff, chief executive of Salesforce, said in a recent podcast that his company also invested in Anthropic as a plan B. According to him, Sam Altman, chief executive of OpenAI, had said that the startup's relationship with Microsoft prevented Salesforce, the software giant's competitor, from investing in OpenAI.
“We thought, ‘Wow, we can’t get in here; we’re going to have to invest everywhere else.’ And we ended up at Anthropic,” Benioff said. "I think we're going to make not just billions, but maybe tens of billions of dollars from this," he added.
Amazon and Google have increased their investments over the years and expanded computing contracts with Anthropic. Amazon is building one of the largest data centers in the world specifically to meet the startup's needs.
After the public offering, few expect the tech giants to quickly sell most of their holdings. Google, Microsoft and a few other companies plan to buy more shares in a private placement parallel to Anthropic's IPO. "Their participation is strategic," said Luria. The relationship guarantees access to large volumes of computing capacity and agreements involving chips.
Whether big tech companies are ultimately the winners will depend on whether the demand for AI justifies the massive investments being made in the sector.
"There is a circularity to these investments, partly driven by the fear of missing out," said Aswath Damodaran, professor of corporate finance and business valuation at New York University's Stern School of Business.
He compared the euphoria around AI to building the biggest factory ever seen, without knowing exactly what it will produce, whether there will be consumers for its products, or how much they will be willing to pay.
"It's a high-risk gamble," he said. "But I don't think anyone wants to be left out of it."
