Semiconductors and Economic Security: A Whole Map from Industry Structure to Export Controls

Hello, I'm Ryuta Hamamoto from TIMEWELL.

Watching the stock market in 2026, you notice that the lead role has clearly changed hands. NVIDIA's market capitalization broke through $5 trillion for the first time in the world in October 2025, and it still sits as the largest company on the planet. In Japan, Kioxia, which had only just listed in December 2024, briefly surpassed Toyota Motor by market cap on June 12, 2026, taking the top spot among domestically listed companies. That is a rise of more than 50x in just a year and a half. The share prices of semiconductor makers have become the thermometer of the global economy.

That said, the reason I wanted to write this series is not the stock rally itself. In the same year, 2026, a string of news with completely different characters broke out around semiconductors. NVIDIA's AI chips were suspected of flowing from Taiwan through Japan and onward to China, and Taiwanese prosecutors opened an investigation. Then in June, something unprecedented happened: AI models themselves became subject to U.S. export controls. Semiconductors are no longer mere electronic components; they have become national security itself.

This article is a map for grasping the whole picture. How does the semiconductor industry actually work? Why has it become the core of economic security? What laws are different countries deploying? I have organized it so that the flow can be followed even without being an expert, and it serves as a bridge to the detailed articles on each theme. By the time you finish reading all of it, the fragments of news should connect into one large picture.

Semiconductors Are Built on a Division of Labor Where "No Single Company Can Make Them"

The first thing I want you to grasp is the fact that the most advanced semiconductors cannot be completed by any single company. The company that designs, the company that manufactures, the company that sells the manufacturing equipment, and the company that supplies the materials are divided up across national borders.

The image might be close to a full-course meal at a fine restaurant. The person who plans the menu, the kitchen that actually does the cooking, the artisan who builds the special oven that only that kitchen can use, the producer who raises the finest ingredients. The roles are cleanly separated, and if even one of them is missing, not a single dish comes out. Semiconductors are exactly like this: no matter how brilliant a designer you have, no single company can make even one grain of an advanced chip on its own.

Following the rough flow, it goes like this. The UK's ARM provides the intellectual property that forms the "foundation" of circuit design, and design-specialist companies like NVIDIA, Apple, and AMD (called fabless companies) use it to design chips. But because they do not own their own factories, they outsource manufacturing to Taiwan's TSMC. And TSMC, in turn, can do nothing to etch the most advanced chips without the EUV lithography machines (machines that print circuits using extreme ultraviolet light) made by the Netherlands' ASML. Furthermore, those machines and factories run on top of the silicon wafers (the disc-shaped material that becomes the foundation of a chip), for which Japanese companies such as Shin-Etsu Chemical and SUMCO hold roughly 60% of the world.

Mapping it onto the cooking analogy from earlier: ARM is the basic form of the recipe, NVIDIA and the others are the chefs who plan the menu, TSMC is the only kitchen in the world that can finish this ultra-difficult dish, ASML is the special oven that can only be placed in that kitchen, and Japan's materials makers are the source that supplies the ingredients. If any one of them stops, the course never comes out. That is exactly why one company's trouble can halt dinner tables around the world.

What matters here is that this chain is not a single straight road; dependence is concentrated on specific companies. Over 90% of the most advanced logic semiconductors are manufactured by TSMC, and ASML holds a 100% global share of EUV lithography machines. HBM, the high-speed memory used for AI, can only be mass-produced by three companies: SK Hynix, Samsung, and Micron. The symbolic moment is that in 2025, NVIDIA overtook Apple to become TSMC's largest customer. It was the historic instant when the lead role of the semiconductor industry shifted from smartphones to AI.

Once you understand this division of labor and the power balance of who depends on whom, you can see why one company's supply stoppage shakes the entire world. We dig into this in detail, with diagrams, in An Introduction to the Semiconductor Industry Structure and A Correlation Map of the Major Companies.

Why Semiconductors Became a "National Lifeline"

Concentrated dependence breeds efficiency in peacetime, but it becomes a fatal weakness in a crisis. The place where that weakness appears most sharply is Taiwan.

About 90% of the world's most advanced logic semiconductors are made in Taiwan. This single point makes a "Taiwan contingency" the greatest risk to the global economy. If Taiwan's semiconductor supply stops, we can no longer make AI, smartphones, cars, or weapons. Bloomberg Economics estimates that a full-scale conflict over Taiwan would inflict roughly $10.6 trillion in losses on the world economy, about 10% of global GDP. Conversely, the very fact that "the world depends on Taiwanese chips" becomes a deterrent that discourages military invasion. Imagine a situation where everyone in town draws water for their lives from a single well. Destroy that well, and everyone goes thirsty, including the person who destroyed it. So no one can lay a hand on it. Taiwan's semiconductors are the same: attack them, and even the attacker's own economy dries up. This mutual dependence, this "destroy it and you hurt yourself too," is precisely the idea called the "silicon shield."

Cracks are now beginning to form in that shield. In May 2026, Taiwan's Keelung District Prosecutors Office detained three people on suspicion of attempting to divert about 50 AI servers loaded with NVIDIA chips, worth over $15 million, to China by falsely declaring them on paper as bound for Japan. At least one shipment reportedly actually passed through Japan and reached Hong Kong. What I do not want you to misunderstand here is that Japan was not the perpetrator of the smuggling. Japan was misused as a "legitimate destination" in the declaration, in effect a transit point that was exploited, and Taiwanese authorities are at the stage of scrutinizing the customs records. Even so, it was an incident that confronted us with the reality of the most advanced chips the world is trying to fence off slipping through the regulatory net and flowing out.

The stock rally is also two sides of the same coin as this tension. Precisely because semiconductors have become a "strategic resource to be fought over," money is pouring into the companies that hold them. Why a Taiwan contingency and a stock rally are two sides of the same coin is something I write about in detail in The Geopolitics of Semiconductors.

Regulation Itself Became a "Weapon"

Once semiconductors become a strategic resource, countries begin to use laws as weapons. The moves from 2025 into 2026 were exactly that exchange of blows.

When you hear "export controls," it sounds difficult, but in essence it is doing airport baggage screening on a national scale. The state checks "Is this item allowed to be taken out?" and "Is this counterparty allowed to receive it?" and stops anything dangerous. It should be easier to picture if you think of semiconductors and their manufacturing technology as now being on the same "take-out caution list" as handguns and knives. The problem is that this list keeps getting rewritten according to each country's calculations.

The United States, through the Commerce Department's BIS, has used its export administration regulations to progressively tighten exports of advanced semiconductors and manufacturing equipment to China. But from 2025 onward the character changed, and regulation has become a "bargaining card." The global AI-chip allocation rule (the AI Diffusion Rule) created by the Biden administration was rescinded by the Trump administration just before it was to take effect, and NVIDIA's H200, whose exports to China had been blocked, was unblocked on the condition that 25% of the sales be paid to the U.S. government. Loosening and tightening regulation itself has become a tool of dealmaking.

China, for its part, is using mineral resources as its trump card. It put gallium, germanium, and medium-to-heavy rare earths, which are indispensable for semiconductors and rare earths, on a license-required basis, and in October 2025 it even rolled out an extraterritorial rule stating that "even foreign products containing as little as 0.1% Chinese rare earths require Chinese permission." Furthermore, in January 2026, as retaliation for Prime Minister Takaichi's remarks regarding Taiwan, it banned exports of dual-use goods to Japan's defense-related companies. For Japan, this is not someone else's fire across the river.

Japan, too, in 2023 placed 23 categories of semiconductor manufacturing equipment under a METI license requirement, and Taiwan created a new "economic espionage offense" that imposes up to 12 years in prison for technology leaks. Which laws each country uses to protect what is something we organize country by country in Each Country's High-Tech Regulations.

AI Explodes Demand, and the Front Lines Move to "Power" and "Space"

To the story of regulation and geopolitics, one more enormous pressure is added. AI.

The evolution of generative AI has pushed semiconductor demand into a different dimension. As of October 2025, NVIDIA publicly disclosed that it was holding $500 billion worth of orders, and just four companies, Microsoft, Google, Amazon, and Meta, are expected to make data-center investments totaling over $700 billion in 2026. But here an unexpected bottleneck appeared. Power. You buy thousands of the latest sports cars, but there is not a single gas station nearby where you can refuel. What is happening now is exactly that situation. No matter how many high-performance GPUs you line up, without the electricity to run them, they are just boxes. In the U.S., data-center power demand has surged, and the capacity of the transmission grid can no longer keep up. That is precisely why Microsoft restarted the Three Mile Island nuclear plant, and companies are racing to secure nuclear power and SMRs (small modular reactors, next-generation reactors that can be built smaller than conventional ones). What came after the semiconductor shortage was a power shortage.

And there is even a move to solve that constraint "off the planet." SpaceX, which only just listed on Nasdaq on June 12, 2026 (its roughly $75 billion raise is the largest IPO in history), unveiled a concept for an orbital data-center satellite, "AI1," carrying AI computation. Google is advancing "Project Suncatcher," in which it launches its own TPUs into space, and Starcloud, in which NVIDIA has invested, was already running H100s in space in November 2025. In space there is unlimited sunlight and no need to worry about cooling. It may sound far-fetched, but it is the flip side of just how serious the power constraint has become. How AI, semiconductors, power, and space connect is something we cover in How AI Is Reshaping Semiconductors, Power, and Data Centers.

At Last, "AI Models Themselves" Became Subject to Export Controls

Once you take the flow this far into account, the meaning of what happened in June 2026 comes into view. The object of export controls jumped from chips to "AI models themselves."

On June 12, 2026, the U.S. Commerce Department's BIS, acting on an order from Commerce Secretary Lutnick, suspended all foreign access to Anthropic's top-tier AI models, "Fable 5" and "Mythos 5." Let me write this carefully: classifying the weights (parameters) of AI models as subject to regulation had already begun with the 2025 AI Diffusion Rule. What makes Fable 5 a "first" is the point that an individual operational order was issued against a specific AI model that is actually deployed and used by hundreds of millions of people. Until now, what had been controlled were chips, manufacturing equipment, and computing resources. This time, a finished AI service was directly stopped.

Why go this far? If the regulation up to now was baggage screening that stops the "weapon itself" at the border, this time the difference is that they went after the "tutor who teaches you, step by step, how to build the weapon." Behind it is wariness of the "weaponization" of AI. The most advanced AI, called frontier AI, can, depending on how you ask, "generate" technical information that is supposed to be tightly controlled, such as missile guidance algorithms, semiconductor manufacturing techniques, and methods of cyberattack. Unlike a search engine that retrieves existing information, it stitches fragments together and writes a new blueprint, and that is what does not fit within the conventional framework of regulation. This order, too, was reportedly triggered by reports of techniques that could bypass safety mechanisms and have the model identify vulnerabilities in critical infrastructure. On the other hand, Anthropic strongly objects to this order, and negotiations toward restoration are ongoing. As of June 2026, it is still an unresolved and contentious matter.

Even so, I believe this trend will not stop. Export controls have climbed up one step at a time, from semiconductors to manufacturing equipment, to computing resources, and now to AI models. The higher AI's capabilities rise, the more regulation will turn toward the contents of the models themselves. For details, we dig into the Fable 5 case and the weaponization of AI in The Era When AI Models Become Subject to Export Controls.

So, What Should Companies Do?

Having read this far, some of you may feel, "It's a grand story, but what does it have to do with my own company?" If anything, it is the opposite: this change is directly tied to the practical work of ordinary companies.

Of course manufacturers and trading companies that handle advanced semiconductors and manufacturing equipment are affected, but companies that have built AI into their operations cannot stay uninvolved either. Before exporting or providing something, you need to perform a "classification screening" to determine whether that product or technology is subject to regulation, and you are also expected to screen whether your counterparties are restricted-entity companies. Neglect the screening and you face severe penalties as a violation of the Foreign Exchange Act, and the rules change not quarter by quarter but on the order of weeks. As with the Fable 5 case mentioned at the outset, even a situation where AI that was usable yesterday is stopped today has become a reality. It is no longer a speed that human hands alone can keep up with.

This is where TRAFEED, which we are developing, comes in. TRAFEED is an export-control AI agent compliant with the standards of Japan's Ministry of Economy, Trade and Industry; it automates classification screening and counterparty screening and keeps up with regulations that change moment by moment. It is an ironic story, but in an era when AI becomes subject to regulation, the one handling that regulation is, again, AI. How to design the practical work of export control is something we explain concretely in Export Compliance Practice for Companies.

The story around semiconductors begins with the industry structure and then connects, all in one line, to geopolitics, the laws of each country, AI and power, and the regulation of AI models. No matter which piece of news comes through next, if you keep this map at hand, you should be able to tell where in the whole picture it is happening. Let's move on to the specifics and raise the resolution.

The primary sources and reporting I referred to are compiled at the end of each detailed article.