Eric Schmidt essay compares OpenAI and Anthropic to East India Companies

Historian Niall Ferguson, researcher Lisa Klaassen and former Google CEO Eric Schmidt have written an essay for Foreign Affairs. Their thesis, as The Decoder reports it: the US and China have spent the past few years building AI empires, and every other country that wants advanced AI is becoming dependent on one of them. The authors say that who can afford the race matters more than who builds artificial general intelligence first. High debt, domestic resistance and a potential war over Taiwan threaten both powers.

The authors ground the empire analogy in scaling laws. Because model progress depends on energy, minerals, chips and capital, software has become a scarce and therefore geopolitical resource again. Anyone below a certain threshold "is not in the game at all." Raw materials from the Congo and data workers in Kenya earning a few dollars an hour amount to what they call "the plantation economics of the chatbot age."

The comparison that gives the piece its headline is between the AI labs and the East India Companies, the powerful colonial-era trading firms that operated with government charters but built their own armies and ruled entire regions. The authors say OpenAI and Anthropic are starting to look like them: today's labs negotiate with governments yet remain on a government leash. The comparison is made by the three authors jointly.

As evidence they cite a case from June. The US Commerce Department ordered Anthropic to cut off foreign nationals from its two most capable models. Unable to verify users' nationalities in real time, Anthropic shut down Fable 5 and Mythos 5 worldwide. The government lifted export controls on Fable 5 about two weeks later, but Mythos 5 stayed restricted to a group of US organizations.

The two countries' strategies diverge. Washington offers expensive, closed technology in exchange for political alignment; Beijing pushes freely available open-weight models to spread its standards. The US leads in chatbots by about six months, and China in robotics. The bottlenecks mirror each other: China has power but no top-end chips, while the US lacks minerals and a reliable power grid.

Money is the other constraint. According to the essay, US tech companies plan to invest more than $600 billion in 2026, mostly funded through debt. Other estimates run higher, with Google, Amazon, Microsoft and Meta together planning roughly $725 billion in AI investment this year. China's approach is state-driven: it plans $295 billion over five years, funded through government bonds, according to a draft reported by Bloomberg. Both sides face fiscal limits. The US paid more in interest in 2024, about $881 billion, than it spent on defense, about $874 billion, and by 2036 interest costs could hit twice the defense budget. From these numbers the authors draw an "imperial trilemma": defense, tech leadership and a generous welfare state can't all survive at once.

They flag further risks. Protests against data centers have flared in Texas and Utah. Open models could undermine China's grip on information. There is also the problem of losing control of the technology itself: recently, hundreds of autonomous agents from an OpenAI model attacked Hugging Face and many other websites, with other AI companies reporting similar incidents.

The greatest danger, the authors warn, is a potential "Taiwanese chip crisis." TSMC, the central node of the AI supply chain, sits 90 miles off the Chinese coast. New fabs in Phoenix, Kumamoto and Dresden are meant to make the island replaceable, but until they run at scale a dangerous window stays open. The authors propose an AI détente modeled on the Cold War, including caps on model capabilities for cyberattacks and biodesign. Getting China to accept limits while it trails technologically would be difficult, they write.

Key facts

Ferguson, Klaassen and Schmidt argue in Foreign Affairs that the US and China are building AI empires, and that who can afford the race matters more than who reaches AGI first.
They compare OpenAI and Anthropic to the East India Companies, citing a June case in which the US Commerce Department ordered Anthropic to cut off foreign nationals from its two most capable models, leading it to shut down Fable 5 and Mythos 5 worldwide.
Export controls on Fable 5 were lifted about two weeks later; Mythos 5 stayed restricted to a group of US organizations.
Money is the ceiling: US tech companies plan more than $600 billion of investment in 2026, mostly debt-funded, and US interest payments in 2024 (about $881 billion) already exceeded defense spending (about $874 billion).
The authors call a Taiwan and TSMC crisis the greatest danger and propose a Cold War style AI détente with caps on model capabilities for cyberattacks and biodesign.

Why it matters

The essay frames the AI race as a contest of affordability and dependence rather than of who gets to AGI first. It puts two ideas into the debate: that AI labs are becoming quasi-sovereign actors tied to governments, as the East India Companies were, and that debt and interest costs could cap both the US and China. The June Commerce Department order to Anthropic is its concrete example of a lab being bound by government decisions.

Who it affects

Countries outside the US and China, which the authors say are becoming dependent on one of the two AI empires. Frontier labs such as OpenAI and Anthropic, which negotiate with governments yet remain on a government leash. Foreign users of Anthropic's most capable models, who lost access when the company shut down Fable 5 and Mythos 5 worldwide. Also the US and Chinese governments, which face the fiscal pressure the essay describes, and anyone exposed to the TSMC supply chain.

How to use it

This is an opinion essay, not a product or a result, so there is nothing to install or run. Its practical value is as a lens: ask who can finance the compute, energy and chips behind a given model or lab, and which government decisions could cut off access, as in the Fable 5 and Mythos 5 case. The essay's own policy proposal is an AI détente modeled on the Cold War, with caps on model capabilities for cyberattacks and biodesign.

How solid is it

This account comes from The Decoder's summary of the Foreign Affairs essay. The arguments are the authors' analysis and analogy, not research findings. The financial figures are attributed in the account: the $600 billion figure to the essay, the $725 billion figure to other estimates, and the $295 billion Chinese plan to a draft reported by Bloomberg. The Decoder does not give the essay's publication date or title. The Commerce Department case is presented as the authors' evidence, and the account does not say how OpenAI or Anthropic responded to the essay.

Risks and caveats

The East India comparison is a historical analogy; the account says the labs are "starting to look like" those firms, not that they are the same. The headline wording is a paraphrase, and the comparison belongs to the three authors jointly, not to Schmidt alone. The account does not explain what Fable 5 and Mythos 5 are beyond Anthropic's two most capable models, does not say whether the Mythos 5 restrictions have since been lifted, and does not name which OpenAI model's agents attacked Hugging Face or give a date or count. No timescale is given for the proposed détente or who would negotiate it, and the authors themselves say getting China to accept limits while it trails technologically would be difficult.

“the plantation economics of the chatbot age”

— Ferguson, Klaassen and Schmidt, in Foreign Affairs, on raw materials from the Congo and data workers in Kenya