Vietnam's Asymmetric Hedging: Strategic Implications of Establishing an AI Regulatory Framework and Pursuing Semiconductor Self-Reliance

Executive Summary

Vietnam is pursuing semiconductor self-reliance focused on narrow application areas, preparing a list of 20 specialized chip groups through its National Centre for Semiconductor Chip Pilot Production Support. At the same time, it has codified the first risk-based AI classification system in ASEAN through its AI Act and Decree No. 142/2026/ND-CP. The core of the issue is an asymmetric hedging structure: while institutionally leaning toward China, for instance by joining WAICO, Vietnam’s technology stack relies overwhelmingly on investment from Nvidia and U.S.-developed open-weight models. This reflects a "low-intensity bandwagoning" strategy—akin to signing an agreement without being a founding member—calculated to minimize backlash from Washington. However, this is the very point that would be the first to fracture should U.S. pressure for a binary choice materialize. The most likely scenario is this ambiguity persisting for over 12 months. South Korea should focus on exposure management by regularly tracking the enforcement of the AI Act and the finalization of the specialized chip procurement list, rather than attempting to alter Vietnam's political choices. Securing tangible benefits will hinge on creating opportunities for South Korean companies to respond early to regional norm-setting in ASEAN and to evolving trade and economic security dynamics.

I. Situational Analysis

Vietnam's Establishment of an AI Regulatory Framework and Pursuit of Semiconductor Self-Reliance: A Situational Analysis

1. Background and Developments

Vietnam's AI and semiconductor policies are not simply industrial promotion measures from a single ministry. They are the product of a structure in which the Ministry of Science and Technology simultaneously oversees both regulation and industrial development. The enactment of the AI Act and the establishment of the National Centre for Semiconductor Chip Pilot Production Support were pursued in parallel under this ministry.

The core of the AI Act is its classification system for high-risk AI systems. The law defines high-risk AI as "systems that can cause significant harm to human life and health, the legitimate rights and interests of organizations or individuals, national interests, public interests, or national security" [4]. The classification criteria include the impact on human rights, safety, and security; the field of use; the scope of users; and the scale of impact [4]. The subsequent implementing decree, No. 142/2026/ND-CP, specifies the degree of automation, the system's role in final decision-making, and the potential for human oversight and intervention as additional assessment factors [4]. Although this framework resembles the risk-based approach of the EU AI Act, its prominent inclusion of national security requirements reflects the control priorities of Vietnam's communist regime.

Developments in semiconductor policy were further detailed at a seminar in Hanoi on September 10. Nguyen Anh Tuan, director of the National Centre for Semiconductor Chip Pilot Production Support, announced that a list of approximately 20 specialized chip groups is being prepared as priorities for national procurement [1]. The target areas include AI, edge computing, next-generation communications, sensors, the Internet of Things, power electronics, and hardware security [1]. The policy is notable for its narrow focus on developing chips for specific applications rather than mass-producing general-purpose semiconductors.

2. Current Situation

Local media in Vietnam are portraying the new AI regulatory framework as an opportunity for the country to secure a leading position in Southeast Asia. The Vietnam Investment Review, for instance, reported on the AI Act and its decree under the headline, "Vietnam can get ahead in AI regulatory framework" [4]. With many ASEAN member states yet to enact comprehensive AI legislation, the pace of Vietnam's lawmaking can be interpreted as an attempt to gain influence in regional norm-setting. This is a significant development in the context of political and economic trends within ASEAN. By codifying the region's first risk-based AI classification system, Vietnam appears to be positioning itself as a key proposer of norms in discussions on ASEAN digital integration.

The semiconductor policy is similarly designed to use national procurement as leverage to attract private investment. The plan is that once the list of 20 specialized chip groups is finalized, these items will be given priority in government procurement. This appears to be a strategy to lower the initial investment risk for private firms by having the state create the initial demand [1]. However, Director Tuan's comments at the seminar indicated that the list remains in the preparatory stage, with no specific implementation timeline or budget yet disclosed.

Compared to the push for semiconductor and AI self-reliance in other regional countries, Vietnam's moves appear to be lagging. China's latest five-year plan designates full-process self-sufficiency in semiconductors as a top priority [9], aiming for industry revenues to surpass 30 trillion yuan by 2030 [8]. India is also developing its own AI semiconductors, investing $15.4 billion in its Semicon 2.0 program [5]. Vietnam's specialized chip strategy, unlike these efforts, focuses on niche areas rather than full-process self-reliance, a pragmatic approach that reflects its resource constraints.

3. Key Actors and Positions

The Ministry of Science and Technology is the key actor overseeing both the enactment of the AI Act and its decree and the creation of the specialized chip procurement list. The fact that a single ministry handles the potentially conflicting functions of regulation and industrial promotion raises concerns about the lack of an adjustment mechanism should high-risk AI regulations clash with national goals for developing the semiconductor and AI industries.

The National Centre for Semiconductor Chip Pilot Production Support is the executive body for specialized chip development. Director Tuan's remarks indicate the government's intent to generate demand through procurement during the initial phase of market formation [1]. The Centre's role is more akin to an industrial support agency than a regulator, which creates potential for practical tension between its mission to develop specialized chips for defense and security (which could be classified as high-risk under the AI Act) and the need for regulatory compliance.

ASEAN Member States are currently observing Vietnam's enactment of its AI Act. With no unified AI norm at the ASEAN level, Vietnam's legislation could potentially serve as a de facto reference for a regional standard. However, this is an assessment made by local Vietnamese media [4], and it remains unconfirmed whether other ASEAN members will actually follow its example.

The United States and China are not direct actors in this specific policy development, but they form the structural context. U.S. export controls on semiconductors and AI chips to China, based on the security logic that advanced technology is a dual-use asset, are reshaping supply chains [6]. China is countering this by aiming to internalize the entire production process [6][9]. Vietnam's specialized chip strategy can be viewed as an attempt to carve out an independent niche amid the technological standards competition between these two powers. However, it is still unclear which side's technology stack the 20 proposed chip groups will ultimately depend on.

4. Key Issues

The first key issue is the feasibility of simultaneously pursuing regulation and industrial promotion. The AI Act's criteria for high-risk classification broadly encompass national security and public interest [4], which means that AI and hardware security projects on the specialized chip list could potentially fall under its regulatory scope. The mechanism for coordinating the pace of industrial development with the intensity of regulatory enforcement remains unclear.

The second key issue is supply chain dependency. Although the strategy for the 20 specialized chip groups aims to create initial demand through national procurement, reliance on foreign technology for design, manufacturing equipment, and materials is inevitable. From a trade and economic security standpoint, this could lead to a shifting of dependencies on critical items rather than a fundamental resolution of the risk.

The third key issue is the sustainability of Vietnam's position as a normative leader within ASEAN. It is confirmed that Vietnam has enacted the region's first risk-based AI legislation [4]. However, based on the available data, it is difficult to assess how consistently this framework will be enforced or how much influence it will exert on lawmaking in other ASEAN member states. The gap between the enactment of the law and its actual diffusion as a regional norm will be a critical focus for future observation.

II. In-Depth Analysis

Vietnam's Establishment of an AI Regulatory Framework and Pursuit of Semiconductor Self-Reliance: An In-Depth Analysis

1. Analysis of Root Causes

Vietnam's enactment of its AI Act and its pursuit of semiconductor self-reliance are not separate issues; both stem from the same structural anxiety. The country's leadership is making a sober assessment of the vulnerable position a nation faces in an era of great power competition when it lacks control over technological standards and supply chains.

It is no coincidence that the Ministry of Science and Technology oversees both the AI Act and the semiconductor support center. This indicates an intent to handle regulatory and industrial sovereignty as a single policy package. While the high-risk AI classification system borrows the EU AI Act's risk-based framework, its emphasis on national security reflects the distinct control priorities of Vietnam's communist regime [4]. This approach differs from the European discourse centered on protecting personal data and human rights. Vietnam's rationale for regulation appears driven more by a determination for the state to maintain control over the pace and direction of AI's spread than by a primary focus on citizen protection.

The fundamental driver of the specialized chip policy is more direct. The list of 20 specialized chip groups being prepared by the National Centre for Semiconductor Chip Pilot Production Support focuses on applications with military and security implications, including AI, edge computing, next-generation communications, and hardware security [1]. Lacking the capital and technology to compete in the general-purpose foundry market, Vietnam has chosen a strategy of using state-led procurement in a narrow niche to stimulate private investment [1]. This approach is less about achieving full self-reliance and more about an attempt to "reconfigure dependency to a manageable level."

2. Structural Context

Political Structure: The Logic of Technological Control in a One-Party System

In Vietnam's one-party system, regulating new technology always serves the dual purpose of industrial promotion and the maintenance of political stability. The inclusion of "significant harm to national interests, public interests, or national security" in the AI Act's high-risk criteria [4] can be interpreted as a measure to retain political control over generative AI services that could shape public opinion. In this regard, Vietnam's approach to AI regulation has more in common with China's model of cyberspace management than with the EU's model of exporting market-based norms.

Economic Structure: An Attempt to Leap from an Assembly-Based Economy to Design Capabilities

For the past two decades, Vietnam's economy has grown by serving as an assembly hub for multinational electronics firms. Attracting the final assembly operations of foreign companies, such as Samsung Electronics, has been a cornerstone of this growth. However, this assembly-centric model is characterized by low value-added and is vulnerable to supply chain realignments amid intensifying U.S.-China technological competition. The list of 20 specialized chips can be viewed as a first step for Vietnam to internalize some design and production capabilities and move beyond its role as an assembly base [1]. Nevertheless, this move does not fundamentally resolve the risks of dependency on critical goods from a trade and economic security perspective, as the equipment, materials, and design tools required for specialized chip production will still rely on firms from the United States, Japan, and the Netherlands.

Security Structure: Reaffirming its Position Between the U.S. and China

Vietnam's pursuit of semiconductor self-reliance is an indirect but clear consequence of the U.S.-China strategic competition. With China elevating full-process semiconductor internalization to a top national priority in its 15th Five-Year Plan [9][8] and the United States extending its export controls on China to advanced equipment and materials [6], Vietnam is attempting to mitigate the risk of being exclusively integrated into either supply chain. Vietnam's specialized chip policy, however, charts a third course: it seeks practical self-reliance in niche applications without being fully incorporated into either China's internalization strategy or the U.S. friend-shoring strategy.

3. Comparison with Historical Precedents and Similar Cases

Structural Similarities with India's Attempt at Semiconductor Self-Reliance

India's effort to develop indigenous AI semiconductors offers a valuable point of comparison. The rationale for India's technological self-reliance dates back to 1985, when the United States denied the sale of a supercomputer over concerns about India's nuclear program. This experience became a formative event for Indian policy circles, cementing the belief that dependence on foreign technology directly equates to security vulnerability. The Modi government has continued this legacy, investing $15.4 billion in its Semicon 2.0 program as part of a third-country hedging strategy aimed at resisting pressure to align with either the U.S. or Chinese AI platforms [5].

The crucial difference between Vietnam and India lies in scale and financial resources. India has the capacity to independently fund a $15.4 billion national program. Vietnam, in contrast, must pursue similar objectives using a narrow list of 20 specialized chip groups and the limited leverage of national procurement [1]. As the Indian case has already demonstrated, the semiconductor industry's need for long-term capital and technological accumulation means there is a high probability that a significant gap between policy announcements and actual mass-production capabilities will persist for some time [5]. Vietnam is not immune to this challenge.

Contrast with China's Full-Process Internalization Strategy

China's approach operates on a scale completely opposite to Vietnam's. Its latest five-year plan designates a "full-process breakthrough" in integrated circuits as the top priority for the electronic information manufacturing sector. This is assessed as a strategic pivot from merely addressing component supply issues to fundamentally restructuring the entire semiconductor industry [9]. President Xi Jinping has also repeatedly stressed supply chain resilience and self-reliance at national conferences on advanced manufacturing [14]. China aims for industry revenues to surpass 30 trillion yuan by 2030, seeking to build a self-contained national industrial ecosystem [8].

Vietnam's specialized chip strategy is an initiative on a completely different plane than China's full-process internalization. Vietnam is not attempting to build a self-contained semiconductor ecosystem. Instead, it is adopting a selective approach aimed at securing a partial stake in the design and production for specific high-growth applications, such as AI and edge computing [1]. While this is a pragmatic choice that recognizes its resource and technology gaps, it is also a tacit acknowledgment of the structural reality that Vietnam cannot be fully autonomous from the semiconductor ecosystems of either the United States or China.

Attempting to Preemptively Set Norms in ASEAN: An Independent Path, Not a Derivative of the EU AI Act's Diffusion

Vietnam's AI Act is the first codified regulation of its kind in a region where most ASEAN member states still lack comprehensive AI legislation. The local media's portrayal of this as a chance for "Vietnam to get ahead in AI regulatory framework" [4] is more than self-congratulation; it reflects a political calculation to claim the role of a norm-setter in discussions on ASEAN digital integration. This effort, however, is not a simple copy of the EU's model of exporting regulatory norms. The framework's prominent inclusion of national security provisions suggests that its primary focus may be on securing state control for individual nations rather than on promoting regulatory harmonization across the bloc.

4. Key Variables Shaping Future Developments

The Timing and Budget for the Finalized Specialized Chip List

The list of 20 specialized chip groups revealed at the September 10 seminar remains in the preparatory stage [1]. The timing of the list's finalization and its integration with national procurement, as well as the size of the allocated budget, will be the first major test of the policy's effectiveness. A significant gap between the announcement and its execution could be seen as a repeat of the policy-capability disconnect already witnessed in the case of India.

Shifts in the U.S.-China Export Control Frontier

Vietnam's specialized chip strategy is directly exposed to shifts in the export control and industrial policies of both the United States and China. The U.S. expansion of measures to third countries like Mexico to prevent the circumvention of export controls on AI hardware to China [13] suggests that similar pressure could be exerted on Vietnam. Depending on which supply chains Vietnam relies on for the design tools and equipment necessary for chip development, its pursuit of self-reliance could itself become ensnared in the regulatory web of one of the major blocs.

The Diffusion of Norms within ASEAN

Whether other ASEAN member states adopt Vietnam's AI Act as a legislative model will be a key indicator of its ability to secure genuine normative leadership in the region. At present, there is no evidence that other countries in the region are referencing Vietnam's legislation. The future of this variable will depend on the pace of discussions among ASEAN members regarding the coordination of digital norms.

The Maturity of the Domestic Workforce and Infrastructure Base

For both specialized chip development and the enforcement of AI regulations, Vietnam's ultimate constraint is an absolute lack of skilled personnel and computing infrastructure. The fact that the National Centre for Semiconductor Chip Pilot Production Support is limited to supporting pilot production [1] indicates that Vietnam's semiconductor industry remains in an experimental, pre-mass-production stage. How quickly this gap can be bridged will be the most fundamental variable determining the policy's success or failure.