Global technology shares fell on Tuesday, November 17, as investors assessed new United States restrictions on advanced artificial-intelligence chips and prepared for Nvidia’s latest quarterly results. The market reaction has put the economics of overseas data centres, access to high-performance computing and the durability of AI-related investment back in focus.

NVIDIA graphics processing unit chip on a circuit board
Photo: Mickael Courtiade. Source · CC BY 2.0.

Asian technology shares lead the decline

Reuters reported that Asian equities weakened as investors reacted to Washington’s move to tighten controls on advanced computing chips and to uncertainty surrounding the next stage of AI infrastructure spending. The Financial Times reported that South Korea’s Kospi fell 2.6 per cent, Taiwan’s Taiex lost 2.3 per cent and Japan’s Topix declined 1.6 per cent. European shares also moved lower, with the Stoxx Europe 600 down 0.7 per cent in the session described by the newspaper.

The pressure extended beyond chip designers. Semiconductor manufacturers, equipment suppliers and data-centre companies are linked through a supply chain that depends on access to advanced processors, power, cooling systems, networking equipment and large pools of capital. When rules change in one part of that chain, investors reassess the costs and expected returns of projects elsewhere.

Reuters also reported declines in Chinese internet shares, including Alibaba and Baidu, as investors considered the effect of tighter access to advanced computing resources. Market moves were not uniform across all assets: the same Reuters report noted weakness in some commodity prices, a firmer US dollar and a modest rise in gold.

Processed semiconductor wafer containing hundreds of prospective chips
Photo: Phiarc. Source · CC BY-SA 4.0.

What the new US restrictions cover

The US Commerce Department’s Bureau of Industry and Security has issued an interim final rule aimed at limiting remote access to advanced American computing chips and servers through data centres located in third countries. The measure is designed to prevent companies based in restricted destinations from obtaining significant computing capacity indirectly by renting services from overseas facilities.

The rule identifies Indonesia, Malaysia, Singapore, Brazil and Thailand for additional destination-based licensing requirements on specified advanced computing items. It also adds end-use controls for certain remote computing arrangements. The Federal Register document explains that the policy is intended to address the possibility that advanced chips could be installed in a third country and then used remotely by organisations that would face restrictions if they attempted to obtain the hardware directly.

The Bureau of Industry and Security says the rule is intended to protect national security while allowing legitimate commercial use of US technology. Its detailed text includes licensing procedures and changes to earlier regulatory provisions. The practical impact will depend on how the requirements are implemented, the scope of exemptions and the compliance systems adopted by chip suppliers and data-centre operators.

Silicon wafer used in semiconductor manufacturing
Photo: Inductiveload. Source · Public domain, released by the creator.

Why Southeast Asia is exposed

Several Southeast Asian markets have attracted investment in data centres because of their connectivity, growing digital economies, access to regional customers and proximity to major technology markets. Reuters reported that the new controls threaten a regional AI and data-centre industry valued at more than $100 billion, with companies involved in data centres and chip supply chains facing a reassessment of their business models.

The concern is not simply whether a facility can purchase a processor. Operators may also need to establish who can access the computing capacity, how much capacity is supplied to a particular customer, where the customer is headquartered and whether the arrangement triggers a licence requirement. That can add compliance costs and uncertainty to projects that were planned around cross-border cloud services.

For chip manufacturers, the issue is equally complex. Export controls can narrow the pool of eligible customers or delay sales, but they may also redirect demand towards other markets and product categories. The immediate market reaction reflects uncertainty about those outcomes rather than proof that global demand for AI computing has disappeared.

Nvidia results become the next market test

Nvidia has scheduled its third-quarter fiscal 2027 results announcement for Tuesday, November 17. The company says its prepared financial commentary will be posted at approximately 1:20 p.m. Pacific Time, followed by a conference call at 2 p.m. Pacific Time. The results cover the quarter that ended on October 25, 2026.

Investors will be watching revenue, data-centre sales, margins, forward guidance and management’s assessment of customer demand. The company’s prior outlook for the third quarter projected revenue of about $108 billion, plus or minus 2 per cent, and said it was not assuming data-centre compute revenue from China. That guidance provides context, but the company’s new results and outlook will be the decisive information for the current quarter.

The results are scheduled for later in the US trading day. At the time of this report, the new quarterly figures had not yet been released, so no earnings outcome or market reaction to the results is being presented as confirmed.

HeadlinesXtra Analysis

The latest sell-off shows how quickly AI-related shares can respond when technology demand and geopolitics intersect. Investors are weighing two different questions: whether companies will continue to spend heavily on AI infrastructure, and whether that infrastructure can be deployed and monetised across borders under changing export rules. The market pressure follows the earlier $4.2 trillion AI infrastructure stock sell-off covered by HeadlinesXtra.

Nvidia’s results may provide an important signal about demand, but one company’s earnings will not settle the broader policy debate. Data-centre operators and chip suppliers need clear, predictable rules, while governments will continue to assess the security risks associated with remote access to advanced computing. The most durable businesses are likely to be those that can demonstrate both strong demand and reliable compliance.

For Southeast Asian economies, the challenge is to preserve investment in digital infrastructure while adapting to new licensing requirements. A pause in some projects would not necessarily end the region’s AI ambitions, but higher compliance costs and uncertainty could influence where future facilities are built and which customers they serve.

Sources and Further Reading