Page 99 - ISC PROCEEDINGS 21.4

AI BEYOND EXPERIMENTATION: TOWARDS PRODUCTIVITY ENHANCEMENT
IN VIETNAM'S DIGITAL ECONOMY

3
2
Phan The Cong* , Nguyen Thi Thu Huong , Trinh Tung , Nguyen Thi Thuy Cham 4
1
1 Thuongmai University, Hanoi, Vietnam.
2 Hanoi Open University, Hanoi, Vietnam.
3 Academy of Policy and Development, Hanoi, Vietnam.
4 Academy of Journalism and Communication, Hanoi, Vietnam.
(*E-mail: congpt@tmu.edu.vn)

ABSTRACT

This paper examines the role of AI in enhancing labor productivity and economic
growth within Vietnam's digital economy during 2021-2025, with forward-looking policy
recommendations toward 2030. Despite significant advances in AI deployment and digital
transformation, evidenced by the digital economy's contribution of 14.02% of GDP in
2025, a region-leading daily AI user interaction rate of 81%, and a 78% surge in AI-
integrated application revenues in the first half of 2025, AI remains predominantly
deployed in fragmented pilot modes, yet to translate into measurable, economy-wide
productivity improvements. This gap between high AI adoption and modest productivity
impact reflects the AI productivity paradox theorized by Brynjolfsson and colleagues.
Employing systematic secondary-source synthesis, GPT theory, and TFP accounting, this
study assesses the current landscape, identifies structural constraints, and proposes
evidence-based policy solutions. Findings reveal three primary barrier clusters: digital
skills deficits, fragmented data infrastructure, and an institutional environment
insufficiently developed to scale AI beyond experimentation. The paper proposes a
comprehensive policy framework to realize AI's potential in achieving Vietnam's targets of
at least 10% GDP growth and 8.5% labor productivity growth in 2026.
Keywords: Artificial intelligence, labor productivity, digital economy, digital
transformation, Vietnam

1. Introduction
The global discourse on artificial intelligence has shifted decisively from
technological fascination to economic consequence. As generative AI integrates into
production processes, supply chains, and public administration, a fundamental question
confronts developing economies: can AI move beyond experimental pilots to generate
measurable, sustained gains in Total Factor Productivity (TFP) and labor productivity at
national scale? This question defines the central inquiry of this paper, and it is nowhere
more pressing than in Vietnam, a lower-middle-income economy pursuing an ambitious
transition toward innovation-driven, high-income status by 2045 (World Bank, 2024).
Vietnam presents a compelling and paradoxical case. The country exhibits some of
the world's highest user-level AI adoption rates: 81% of internet users interact with AI
daily, 83% actively upskill in AI domains, and 96% express willingness to share data with AI
agents, all metrics leading Southeast Asia (Google, Temasek, & Bain & Company, 2025).
Revenue from AI-integrated applications surged 78% in the first half of 2025 alone (VNA,
2025). Yet this adoption intensity has not yet translated into commensurate TFP growth:
Vietnam's digital economy contributed only 14.02% of GDP in 2025, well short of the 2030

98