Eric Schmidt, former Google CEO, speaks at a tech conference, comparing OpenAI & Anthropic to colonial trading companies.

Editorial illustration for Eric Schmidt compares OpenAI and Anthropic to colonial trading companies

Eric Schmidt compares OpenAI and Anthropic to colonial...

By AI Daily Post Edited by Brian Petersen, Editor-in-Chief

October 25, 2026 • 3 min read

Niall Ferguson doesn't usually write about chatbots. But the historian has teamed up with researcher Lisa Klaassen and former Google CEO Eric Schmidt for a Foreign Affairs essay that treats OpenAI and Anthropic less like Silicon Valley startups and more like the East India Company. Their argument: the US and China have carved up the global AI market between them, and everyone else is stuck buying access from one empire or the other.

The split isn't subtle. Washington ties its closed AI systems to political loyalty from allies. Beijing hands out open-weight models for free to lock in its technical standards worldwide. Both approaches cost staggering amounts of money, and both governments are now straining under the debt and interest payments that come with building an AI empire from scratch.

Taiwan sits at the center of the risk. TSMC makes the chips that the entire AI industry depends on, and a conflict there could break the supply chain for both superpowers at once. Ferguson, Klaassen and Schmidt argue that whoever can actually afford to keep playing this game matters more, in the end, than who reaches artificial general intelligence first.

The authors compare OpenAI and Anthropic to the East India Companies, the powerful colonial-era trading firms that operated with government charters but built their own armies and ruled entire regions. Like them, today's AI labs negotiate with governments yet remain on a government leash.

Ex-Google CEO Eric Schmidt calls OpenAI and Anthropic modern colonial trading companies - THE DECODER

Why this matters

Schmidt's comparison is worth sitting with because it reframes OpenAI and Anthropic not as neutral research labs but as instruments of a state project, the way the East India Company once was. For developers building on GPT or Claude APIs, that's a reminder that your tooling choices carry geopolitical weight: closed US models come bundled with alignment to Washington's export controls and chip restrictions, while China's open-weight push is a deliberate standards play, not generosity. Founders raising capital should note the "sky-high investment" pressure Schmidt flags.

These companies need enormous capital, energy, and chip access that only a handful of players can supply, which means the economics of frontier AI are structurally tied to state resources, not just VC rounds. Researchers chasing the "six-month" chatbot lead over China should ask what happens when that gap narrows, especially given China's robotics edge and mineral access the US lacks. None of this is abstract strategy talk.

It's a map of who controls compute, who controls standards, and who you're actually aligning with every time you pick an API.