Shares of Samsung Electronics and Taiwan Semiconductor Manufacturing Company (TSMC) declined on November 17, 2026, after the US government announced fresh restrictions aimed at limiting China’s access to advanced artificial intelligence chips and computing infrastructure. The new rules target not only direct chip shipments but also overseas data centres that allow Chinese companies to rent computing capacity and use advanced processors remotely.
The announcement has raised concerns across Asia’s semiconductor industry, where companies are closely connected through chip manufacturing, memory production, equipment supply and data-centre infrastructure. Samsung and TSMC came under pressure as investors assessed the possibility of additional export licences, customer verification requirements and compliance costs. The measures could also affect Southeast Asian countries that have become important destinations for AI infrastructure investment.
Samsung and TSMC Shares Decline After US Announcement
Samsung Electronics and TSMC shares lost ground during Asian trading on Tuesday following the announcement of the new US restrictions.
The selling pressure reflected concerns that stricter controls could complicate the semiconductor supply chain and affect companies serving Chinese customers.
Other technology stocks also declined. South Korea’s technology-heavy Kospi index fell 2.6%, while Japan’s Nikkei 225 declined 1.3%. Taiwan’s Taiex also came under pressure as semiconductor-related shares weakened.
The broader sell-off reflected investor concerns about the future of AI infrastructure spending and the possibility that tighter regulations could slow business activity across the regional technology ecosystem.
| Company or index | Market reaction reported on November 17 |
|---|---|
| Samsung Electronics | Shares declined during Asian trading |
| TSMC | Shares declined during Asian trading |
| SK Hynix | Shares came under selling pressure |
| South Korea’s Kospi | Closed down 2.6% |
| Japan’s Nikkei 225 | Fell 1.3% |
| Taiwan’s Taiex | Closed down 2.3% |
The market reaction was not limited to chip manufacturers. Data-centre operators, semiconductor equipment suppliers and companies providing infrastructure for AI computing also faced pressure.
What Are the New US Restrictions on China’s AI Chip Access?
The Trump administration announced an interim final rule on November 16 to close a route through which Chinese companies could access advanced American computing technology.
Previously, US export restrictions primarily focused on the physical shipment of certain advanced chips and semiconductor equipment to China.
However, Chinese companies could potentially access computing capacity by renting servers located in other countries. These facilities could contain advanced processors manufactured by American companies, including Nvidia.
The new rules seek to address this arrangement by expanding oversight of overseas data centres and remote computing services.
The US government wants to ensure that companies cannot bypass restrictions on advanced technology simply by accessing the hardware through a cloud service or a data centre located outside China.
The measures introduce additional licensing requirements for covered advanced chips and overseas computing infrastructure.
The countries specifically identified under the new restrictions include:
- Singapore
- Malaysia
- Thailand
- Indonesia
- Brazil
Companies in these markets may need approval from the US Department of Commerce’s Bureau of Industry and Security (BIS) before shipping covered advanced chips to relevant facilities.
The rules also introduce additional scrutiny of overseas data centres supplying significant computing capacity to companies headquartered in countries subject to US arms-related restrictions.
The objective is to make it more difficult for restricted Chinese entities to obtain advanced AI computing power through third-country infrastructure.
Why Samsung Foundry Could Face Additional Compliance Requirements
Samsung is one of the world’s major semiconductor manufacturers and operates a foundry business that produces chips designed by other companies.
Unlike companies that manufacture only their own processors, foundries manufacture semiconductors for customers that develop their own chip designs.
These customers are commonly known as fabless semiconductor companies.
Some fabless businesses based in China depend on external foundries to manufacture advanced chips. This creates additional regulatory responsibilities for manufacturers that serve international customers.
CLSA senior analyst Sanjeev Rana said Samsung Foundry could potentially face additional customer due-diligence and licensing requirements under the new US rules.
The company may need to strengthen its procedures for verifying customers, checking the intended use of chips and establishing whether products could ultimately reach restricted entities.
This could increase administrative costs and make certain manufacturing relationships more complicated.
However, the new rules do not automatically mean that all Samsung chip production for Chinese customers will stop. The impact will depend on the specific chip technology, customer, destination and licensing requirements.
Samsung’s exposure also extends beyond foundry services. The company is a major supplier of memory chips used in servers and AI infrastructure, meaning broader weakness in data-centre investment could affect sentiment towards its semiconductor business.
What Does the US Rule Mean for TSMC?
TSMC is the world’s largest contract semiconductor manufacturer and produces advanced processors for several major technology companies.
Its customers include businesses developing processors for smartphones, high-performance computing and artificial intelligence applications.
The new restrictions could create additional compliance requirements for foundries manufacturing advanced chips for Chinese fabless companies.
However, Reuters did not identify a specific financial impact on TSMC or quantify its exposure to the affected customers.
The company’s global manufacturing business is highly interconnected, making export regulations an important consideration for its customer relationships and production planning.
If licensing requirements become more extensive, foundries may need to spend additional resources verifying customer identities, end users and the final destination of manufactured chips.
This could increase administrative complexity and potentially delay certain orders.
For TSMC, the longer-term concern is whether tighter restrictions will accelerate the separation of semiconductor supply chains between China and the United States.
Such a shift could influence where chip designers develop products, where foundries manufacture them and which markets can legally access advanced computing technology.
Southeast Asia’s $100 Billion AI Infrastructure Industry Faces Pressure
One of the biggest implications of the new rules could be for Southeast Asia’s rapidly expanding data-centre industry.
Countries such as Malaysia, Singapore and Thailand have attracted substantial investments from technology companies looking to establish data centres and expand cloud computing capacity.
These facilities support AI model training, cloud services, enterprise applications and other computing-intensive operations.
However, some Chinese companies have used overseas data centres to access advanced American chips that are difficult to obtain directly under existing export restrictions.
The new rules could therefore affect the business models of data-centre operators serving customers with links to China.
Reuters reported that shares of several Southeast Asian infrastructure companies declined sharply following the announcement.
| Company | Market impact reported on November 17 |
|---|---|
| MN Holdings, Malaysia | Shares fell nearly 20% |
| Critical Holdings, Malaysia | Shares fell nearly 20% |
| YTL Power, Malaysia | Shares declined nearly 10% |
| Frencken, Singapore | Shares fell 3.91% |
| AEM Holdings, Singapore | Shares declined 2.9% |
The market reaction reflects concerns that stricter customer verification could reduce demand from Chinese AI developers and delay data-centre expansion plans.
For operators that depend on international cloud customers, compliance costs could also increase.
Malaysia and Thailand Have Billions of Dollars Invested in Data Centres
The restrictions arrive at a time when Southeast Asian governments are encouraging large investments in digital infrastructure.
Malaysia has approved approximately 385.7 billion ringgit, or around $95 billion, in data-centre-related investments since 2021, according to figures cited by Reuters.
Thailand had 42 data centres representing approximately 3,400 megawatts of IT load and investment of around 750 billion baht, equivalent to approximately $22.6 billion, as of February 2026.
These investments are part of a broader effort to position Southeast Asia as an important destination for cloud computing and AI infrastructure.
However, the new restrictions introduce uncertainty for companies whose facilities may serve customers in restricted markets.
If data-centre operators lose access to some customers or face longer approval processes, utilisation rates and expansion plans could be affected.
The impact will depend on how the US government implements the rules and how operators adapt their customer-verification systems.
China Could Accelerate Its Push for Domestic AI Chips
The restrictions could also strengthen China’s efforts to develop its own semiconductor and AI computing ecosystem.
Chinese technology companies have been working to reduce dependence on advanced American processors by investing in domestic chip design, manufacturing and software infrastructure.
Companies such as Semiconductor Manufacturing International Corporation (SMIC) are important to Beijing’s strategy of building a more self-reliant semiconductor industry.
Following the announcement, SMIC shares rose approximately 7% on Tuesday, according to Reuters reporting.
The movement suggests that some investors see the restrictions as a potential catalyst for greater demand for domestic semiconductor alternatives.
However, developing competitive advanced chips requires more than investment in manufacturing facilities. Companies also need access to sophisticated semiconductor equipment, design software, high-bandwidth memory and advanced packaging technologies.
Restrictions on American technology could make China’s efforts more difficult in the short term while increasing the incentive to develop alternative domestic capabilities over the longer term.
This could accelerate the fragmentation of the global semiconductor industry into technology ecosystems with different suppliers, standards and regulatory requirements.
Nvidia and the Broader AI Industry Face Increased Scrutiny
The new restrictions also have implications for Nvidia, whose advanced processors are widely used in AI data centres.
The company has been at the centre of US-China semiconductor tensions because its products are important for training and operating advanced AI models.
The new rules seek to prevent Chinese companies from accessing restricted computing capacity through overseas facilities.
This could affect demand from data-centre operators that previously served customers with links to China.
At the same time, Nvidia faces a broader investment challenge as investors assess the sustainability of global AI infrastructure spending.
Technology companies have committed enormous amounts of capital to data centres, advanced processors and cloud computing capacity. Any regulatory restriction that makes it harder to deploy this infrastructure could influence investment decisions.
The restrictions therefore come at a sensitive time for the semiconductor industry, with investors also watching Nvidia’s quarterly earnings and the outlook for AI-related capital expenditure.
The Bigger Picture
The latest US restrictions show that the technology competition between Washington and Beijing is expanding beyond the physical export of semiconductor products. By targeting remote access to advanced computing power, the US government is attempting to close a regulatory gap created by cloud computing and international data-centre networks. This could make it more difficult for Chinese companies to access advanced AI infrastructure even when the physical hardware is located outside China.
For semiconductor manufacturers such as Samsung and TSMC, the immediate concern is additional customer verification and licensing requirements. For Southeast Asia, the challenge is balancing its ambitions to attract global data-centre investment with increasingly complex international export-control rules. China, meanwhile, may accelerate efforts to develop domestic alternatives. The result could be a more fragmented global semiconductor industry, with technology access increasingly shaped by national security policies and geopolitical relationships.
Looking Ahead
Investors will monitor how the US Department of Commerce implements the new rules, particularly the licensing requirements for chip shipments and overseas data-centre operations. The impact on Samsung and TSMC will depend on their exposure to affected customers and the extent of additional compliance obligations. Semiconductor equipment suppliers, memory manufacturers and cloud infrastructure companies could also experience changes in demand as customers reassess their operations.
For the global AI industry, the restrictions create another layer of uncertainty around where advanced computing infrastructure can be built and who can access it. Southeast Asian countries will need to clarify how the new requirements affect their data-centre sectors, while Chinese technology companies may increase investment in domestic alternatives. The longer-term consequence could be a restructuring of global AI supply chains, with semiconductor companies having to navigate commercial opportunities alongside increasingly strict geopolitical regulations.
Get the day’s top stories in your inbox
One concise email. No spam, unsubscribe anytime.



