TSMC to Invest Up to $1 Billion in Anthropic and OpenAI as AI Chip Demand Soars

The company that makes the chips behind the artificial intelligence boom is now buying a piece of the boom itself. Taiwan Semiconductor Manufacturing Company (TSMC) has approved investments of up to $1 billion in two of the world’s leading AI developers, Anthropic and OpenAI, in a move that shows how deeply the chipmaker believes in the long-term future of AI — even as worries grow about how much money is being poured into the technology.

TSMC’s board approved the plan on Tuesday, November 10. The company will invest up to $500 million in Anthropic, the maker of the Claude chatbot, buying shares at the price of Anthropic’s upcoming stock market listing. It will invest up to another $500 million in OpenAI, the maker of ChatGPT, buying preferred shares in the company.

It is an unusual step for TSMC. The Taiwanese giant is the world’s largest contract chipmaker: it does not design its own AI systems, it manufactures the chips that other companies design. Now it is becoming an investor in its own customers’ world, betting that the companies building AI will keep growing — and keep needing its chips — for years to come.

What TSMC announced

The details were announced after TSMC’s board meeting in Hsinchu, Taiwan, on November 10. The board approved two separate investments:

• Up to $500 million to buy common shares in Anthropic, PBC, at the per-share price of Anthropic’s initial public offering (IPO).
• Up to $500 million to buy preferred stock in OpenAI Group PBC, the company behind ChatGPT.

The board also approved up to $500 million to set up a new investment company focused on strategic investments in the semiconductor industry and its wider ecosystem. At the same meeting, it approved capital spending of about $40.9 billion to support TSMC’s long-term factory and technology plans.

None of the purchases has closed yet. The Anthropic investment depends on the IPO going ahead, and the OpenAI deal is a private transaction, so its price and exact terms have not been disclosed. But the board’s approval is a clear public signal of intent — and of confidence.

Why Anthropic’s IPO matters

The timing is no accident. Anthropic priced its IPO on the same day, Tuesday, November 10, setting a price range of $70 to $78 per share. At the top of that range, the company would be valued at up to $1.4 trillion — which would make it one of the largest stock market debuts in history.

Anthropic is also raising $26 billion through a separate private share sale, backed by some of the biggest names in technology, including Google, Nvidia, AMD and Microsoft. TSMC is among the buyers in that private sale, alongside Google, Nvidia, Microsoft, AMD, Broadcom and Micron Technology.

The IPO is expected to be priced on November 18, with Anthropic set to list on the Nasdaq stock exchange under the ticker symbol ANT. Until then, the exact size of TSMC’s stake cannot be calculated, because it depends on the final IPO price. What is fixed is the $500 million ceiling.

For TSMC, the Anthropic investment is straightforward in one sense: it is buying shares in a company at the same price as public investors, in one of the most talked-about listings of the year. For Anthropic, having the world’s most important chipmaker as a shareholder is a vote of confidence from the company that literally builds the hardware its AI runs on.

Why OpenAI is a different kind of bet

The OpenAI investment is different. OpenAI is not listed on any stock exchange, so there is no public share price. TSMC is buying preferred stock — a type of share that usually comes with priority over ordinary shares if the company pays out money, but often with fewer voting rights.

No valuation, price per share or closing date has been disclosed for the OpenAI deal. That is normal for a private investment, but it means the public cannot yet judge exactly what TSMC is getting for its money. What is clear is the direction of travel: TSMC wants a financial stake in both of the leading Western AI developers, not just one.

Together, the two investments would give TSMC a foot in both camps of the AI race. Anthropic’s Claude and OpenAI’s ChatGPT are the two most widely used AI assistants in the world, and both companies are spending enormous sums on computing power — computing power that runs, in large part, on chips made by TSMC.

A chipmaker that cannot make chips fast enough

TSMC’s confidence is backed by its own results. Last month, the company reported a sixth consecutive quarter of record profit, and said demand for AI-related chips remained “extremely robust”. It expects full-year 2026 revenue to grow by about 44%, and says demand continues to outstrip its manufacturing capacity — in plain terms, it cannot build chips fast enough to keep up with orders.

TSMC is the world’s largest contract chipmaker and a critical supplier to leading AI firms, including Nvidia, whose expensive graphics processors power many of today’s AI systems. When Nvidia, Google, AMD or Apple need their most advanced chips made, they come to TSMC. That position has made the company one of the most important — and most valuable — firms in the global technology industry.

The company is also expanding production on a huge scale. It is investing $265 billion in manufacturing facilities in Arizona in the United States, and is evaluating potential projects in Texas. The expansion is partly a response to political pressure to make more chips in America, and partly a response to demand that its factories in Taiwan alone can no longer meet.

The “circular investment” worry

Not everyone is cheering. The announcement has revived a debate in the technology industry about “circular investment” — the idea that AI companies, chipmakers and cloud providers are increasingly investing in each other, creating a loop in which money flows around a small circle of companies and makes the whole boom look bigger than it really is.

Large strategic investments of this kind are rare for TSMC, which has not previously taken part in what many see as circular investment in the AI industry. That makes this decision all the more notable — and all the more closely watched.

Liu Pei-chen, a researcher at the Taiwan Institute of Economic Research, told Radio Taiwan International that the move helps TSMC move beyond being a simple wafer foundry and take a bigger share of the AI value chain. By investing in its customers, TSMC can make more flexible and targeted bets in parts of the supply chain such as silicon intellectual property, software, hardware and next-generation AI.

But Liu also warned of the risks. The tech supply chain is shifting from a collection of individual companies to cross-sector alliances bound together by shared investments, she said. That can help companies share the enormous cost of building AI infrastructure — but it also means that if AI does not make as much money as expected, the close financial ties between suppliers and customers could come under pressure.

The trust problem with TSMC’s other customers

There is another, subtler risk. Both Anthropic and OpenAI are actively developing their own chips, in an effort to reduce their dependence on outside suppliers. TSMC’s investment in both companies could affect the trust of its existing customers and raise concerns about fairness in how its most advanced factory capacity is shared out — capacity that is already very tight.

TSMC’s customers include some of the world’s biggest technology companies, and they compete fiercely with each other. If one customer believes TSMC is favouring a company it has invested in, that trust could be damaged. For a company whose entire business is built on being a neutral manufacturer for everyone, that is a real concern.

Regulators are another question mark. Liu noted that governments in many countries are beginning to introduce tighter oversight as global technology competition heats up. Cross-border investments in sensitive technology could face tougher scrutiny in future, which adds further uncertainty to deals of this kind.

Why this matters beyond Taiwan

TSMC’s decision matters for three reasons, and none of them is limited to Taiwan.

First, it is a statement of belief. TSMC knows more about real demand for AI chips than almost anyone, because it sees the orders from every major AI company. When the company that makes the chips says it is confident enough to invest $1 billion in the companies buying them, that carries weight — especially at a moment when investors are asking whether AI spending has gone too far, too fast. In the same week as TSMC’s announcement, Nvidia, the biggest name in AI chips, lost hundreds of billions of dollars in market value in just two days, as investors worried about the scale of AI spending. TSMC’s move pushes in the opposite direction.

Second, it shows how the AI industry is changing. The old model was simple: chipmakers made chips, software companies wrote software, and the two sides did business at arm’s length. The new model is messier and more tangled, with suppliers, customers and rivals all holding stakes in each other. TSMC’s $1 billion is part of that shift — and it is unlikely to be the last deal of its kind.

Third, it underlines how central Taiwan has become to the world economy. A single company’s board meeting in Hsinchu can move markets, shape the future of the world’s two leading AI developers, and draw the attention of regulators from Washington to Beijing. TSMC’s factories are now as strategically important as oil fields were in the last century, and its investment decisions are watched accordingly.

What happens next

The immediate next step is Anthropic’s IPO. The offering is expected to be priced on November 18, and the company will then list on the Nasdaq. Only then will the exact size of TSMC’s Anthropic stake be known.

The OpenAI investment is likely to move more slowly and more quietly, as private deals usually do. There may be no public announcement when it closes, and the terms may never be fully disclosed.

Investors will also be watching TSMC’s own numbers. The company has adopted a cautious outlook for 2027, even as it reports record profits and demand it cannot meet. The question for the year ahead is whether the AI boom keeps growing fast enough to justify the enormous sums being invested in it — by TSMC, by its customers, and by the companies now buying stakes in each other.

If the boom continues, TSMC’s $1 billion could look like a smart, early bet on the two companies shaping the future of AI. If it slows, the same investment could become Exhibit A in the case that the AI industry got too close, too intertwined, and too confident. Either way, the chipmaker that used to stay in the background has now stepped firmly into the spotlight.

Frequently asked questions

What is TSMC?

TSMC, or Taiwan Semiconductor Manufacturing Company, is the world’s largest contract chipmaker. It manufactures chips designed by other companies, including Nvidia, Apple, AMD and Qualcomm. It does not sell chips under its own brand to consumers.

How much is TSMC investing in Anthropic and OpenAI?

Up to $1 billion in total: up to $500 million in Anthropic shares at the IPO price, and up to $500 million in OpenAI preferred stock. The board approved the investments on November 10, 2026, but neither purchase has closed yet.

Why is TSMC investing in AI companies?

The move gives TSMC a bigger share of the AI value chain, beyond simply manufacturing chips. The investments give it a financial stake in two of its most important end customers, and signal confidence that demand for AI computing will keep growing.

What is Anthropic’s IPO worth?

Anthropic has set a price range of $70 to $78 per share, which would value the company at up to $1.4 trillion — one of the largest stock market debuts on record. It is also raising $26 billion in a private share sale backed by Google, Nvidia, AMD, Microsoft and others.

What are the risks of the deal?

Analysts warn about “circular investment” — AI companies and chipmakers investing in each other — and about what happens if AI does not earn as much money as expected. There are also concerns about customer trust, since Anthropic and OpenAI are developing their own chips, and about possible regulatory scrutiny of cross-border technology investments.

Conclusion

TSMC’s $1 billion bet on Anthropic and OpenAI is a small sum next to its $40.9 billion capital spending plan, and a tiny fraction of the trillions flowing into artificial intelligence. But it is a significant moment all the same. The world’s most important chipmaker has stopped being a neutral supplier and started being a shareholder in the AI revolution it helped to build.

Whether that turns out to be a masterstroke or a cautionary tale will depend on something no board meeting can control: whether artificial intelligence lives up to the extraordinary expectations — and the extraordinary spending — now placed upon it.

Sources

• Reuters: https://www.reuters.com/world/china/tsmc-invest-up-1-billion-anthropic-openai-2026-11-10/
• Radio Taiwan International: https://www.rti.org.tw/en/news?uid=3&pid=1050842
• Seoul Economic Daily: https://en.sedaily.com/world/2026/11/11/tsmc-to-invest-up-to-500-million-each-in-anthropic-openai

This article is for general information and education only. It is not legal or financial advice.