Consultations with experts and other studies indicate that AI adoption in Viet Nam is still nascent. In 2021, a Southeast Asia-focused report on AI found about 49 percent of respondents in Viet Nam were still piloting AI initiatives rather than scaling or having achieved full-scale end-to-end implementation in different activities across sectors.i
In January 2021, Viet Nam’s National Strategy for Artificial Intelligence Research, Development and Application through 2030 was issued to realize specific, measurable, and incremental targets by 2025 and 2030.ii
National Strategy on Research, Development, and Application of AI through 2030
- Top five in ASEAN region
- Top four in ASEAN with 10 prestigious AI brands in the region
- Top 60 in the world
- At least one representative in one of ASEAN region’s top 20 AI research and training institutions
- 50 sets of open, linked data in various socio-economic sectors
Institutions and personnel
- Two national innovation centers
- Three national innovation centers + “contingent of high-quality human resources”
- Widespread application in public administration, social services, and urban management
- Universalization of basic AI skills among greater public
- Application in defense, security, as well as humanitarian assistance and disaster response
The country foresees a holistic application of AI, from “building a creative society” and facilitating effective governance, to protecting national security, maintaining social order, and advancing sustainable economic development. The language of the Strategy is one of growth and efficiency, with “a system of legal documents and legal corridors related to AI” supplemented by “complete policies and laws.” While the 14-page document does not make any explicit mention of ethics, there are nonetheless references to “avoiding technology abuse and infringing upon legitimate rights and interests of organizations and individuals.”
In recognizing the value of data as foundational to the efficacy of AI, the Strategy outlines a balance to be struck between two seemingly contradictory imperatives: shared, open, even decentralized databases for the research and development of AI applications, on the one hand, and maintaining the data privacy of individuals and organizations, on the other. The latter aligns with Viet Nam’s constitutionally guaranteed right to privacy and (at the time of writing) ongoing efforts to draft a comprehensive data protection regime.
Usage and Impact
There is in Viet Nam an acute cognizance of a gap in the market for locally- or even regionally-tailored solutions and that innovation must be contextualized and applied for domestic use, in the first instance. In 2019, the Ministry of Information and Communications (MIC) launched “Make in Viet Nam,” an initiative to spur the development of the domestic ICT industry for local solutions which would ultimately boost the country’s global presence. The slogan’s active tense—“make” rather than “made”—was a deliberate choice to promote the spirit of “Vietnamese people proactively, creatively designing, and producing products that contribute to the technology community.”iii Tellingly, the minister, Nguyễn Mạnh Hùng, explained: “We heard and told far more about international stories to Vietnamese people. It is time to tell Vietnamese stories. Every Vietnamese person and each Vietnamese business should join the Make in Viet Nam program and tell their own stories to encourage others to do so.”iv
Use of AI in Biomedical Research
The 1000 Vietnamese Genome Project (1KVG), which began in 2018, is Viet Nam’s first large-scale human genome project by Vingroup Bigdata Institute (VinBigData). Over the past three years, the research team has analyzed the genomes of over 1,000 healthy, biologically unrelated individuals between 35 and 55 years of age, with sufficient phenotypic and demographic information. The individual data was processed through Google, Illumina, and NVIDIA. The genomes of more than 4,000 individuals were also analyzed in relation to common diseases and drug reactions. As a result, “more than 40 million genetic variants, including nearly 2 million common and unique genetic variants of the Vietnamese population were detected.” The project is invaluable for biomedical and genomic research in Viet Nam as it is the first and only dataset characterizing genetic variation for the Vietnamese population. It would also promote medical advancements in precision medicine in Viet Nam.
There exists substantive literature on the ethical considerations that should be considered when conducting whole-genome research. These include the return of research results to participants and their relatives, as well as the maintenance of anonymity throughout the testing and analyzing process. Biomedical research marks a significant step towards AI implementation for public health advances at the national level.
A year earlier, the “Digital Knowledge System of Viet Nam” project was launched on January 1 with a website (https://itrithuc.vn/) serving as an open database for all, especially Vietnamese youth, to contribute to. In addition to government agencies providing information, the site is also meant to archive Vietnamese translations of knowledge resources from around the world to feed into a repository of Big Data that can then facilitate AI technologies and IoT platforms.
Domestic conglomerates such as Viettel and VinGroup have begun to forge ahead in the spirit of “Make in Viet Nam,” spinning off AI research-based institutes and working on machine learning, computer vision, and natural language processing for local use. Viettel’s AI open platform features speech synthesis applications that take into account the regional variances in spoken Vietnamese —northern, central, and southern —for use in announcements, customer service systems, and e-readers for the visually impaired or people on the go. Similarly, its voice-to-note product is supposed to be able to recognize these different regional dialects and convert speech to Vietnamese text directly.v
VNG Corporation’s Zalo Group, Viet Nam’s first technology unicorn, introduced “Kiki” in 2021. Kiki is a Vietnamese virtual personal assistant built in the image of Amazon’s Alexa, Apple’s Siri, Microsoft’s Cortana, or Google Assistant, but specifically adapted to local needs and customs. It is trained to recognize the three Vietnamese regional dialects, and can check the lunar calendar and announce winning lottery numbers.vi Zalo Group’s president, Vương Quang Khải, noted that with Kiki Vietnamese people would no longer be left out of AI’s benefits; an apt reflection of the frustration any non-native English speaker has had to contend with when relying on voice assistants designed and developed by U.S. technology companies. Interestingly, although Google Assistant launched Vietnamese language support in May 2019 for the domestic market, the service appears to have been suspended across its Google Home and Google Nest products after struggling to perform reliably in the country.vii
The pandemic was an impetus for local, more effective, and cost-efficient AI-based solutions for many businesses hit hard by the global economic disruption.viii G-Group, the parent company of Gapo, a local social network with about six million users, initially adopted Meta’s online collaboration software, Workplace, to consolidate and streamline G-Group’s internal communication platforms. However, at USD8 per person per month, the cost of using Workplace became too expensive, prompting G-Group to develop GapoWork. Hà Trung Kiên, chief executive officer of Gapo, attributed the advantages of home-grown platforms to a deeper understanding of Vietnamese culture and business needs as well as lower costs.ix
As Viet Nam’s domestic champions work toward fulfilling local needs, they may eventually scale up to meet underserved regional demands. VinGroup’s VinAI, which already claims to rank among the world’s top 25 AI research companies, alludes to this in describing its work: “Due to our unique location we are also naturally drawn toward problems for developing countries, which might otherwise be overlooked in the research community.”x
Challenges and Prospects
Viet Nam’s deployment of AI technologies across public administration, the economy, and daily social life is not without its challenges. As in other countries, these include competition for funding and skilled personnel—issues that are discussed in great detail elsewhere, and are beyond the focus on ethics in this report.xi
Level of awareness
If the state of AI adoption in Viet Nam is still only in its early stages, the topic of ethics in relation to AI has barely begun to percolate in the country, despite the fervor surrounding its digitalization drive. As outlined above, there is some recognition in the National Strategy that there should be ethical or legal parameters to its execution. The Ministry of Public Security, for example, has been tasked to “develop and complete additional legal documents on privacy protection and human rights, on security and social order issues related to AI development and application activities.”
Otherwise, there appears to be little emphasis on delving deeper into the concept, let alone the practical application, of AI ethics in the national context. In mandating the Ministry of Defense to develop and deploy AI applications as part of its military modernization and combat plans, the National Strategy makes no related mention of ethics or international law. However, one informant offered that there is, in fact, awareness in Viet Nam about the risks of lethal autonomous weapons that are currently being debated in the UN and around the world.xii
Likewise, the Ministry of Public Security’s charge to use AI applications in policing and immigration “to proactively prevent crimes” begs further discussion about the implications of surveillance technologies, algorithmic processing, and quality of datasets for predictive policing purposes. Their potential for efficiency notwithstanding, reports about bias and discrimination resulting in miscarriages of justice in the United States and the United Kingdom provide cautionary lessons for Viet Nam and Southeast Asia at large.xiii
Similarly, although the Ministry of Natural Resources and Environment is to promote the adoption of AI applications for data-gathering on land and environmental resources in response to pollution and climate change, the Strategy is silent on the principles or regulatory oversight needed to ensure that these AI-based technologies support, rather than obstruct, Viet Nam’s SDG commitments.
AI, therefore, is seen by the Vietnamese government and public as a generally positive and catalytic force for practicality, convenience, and development. The impact and trade-offs of AI algorithms, however, do not yet appear to be widely considered. For example, AI-powered “rice ATMs”, or automatic rice dispensing machines, provided free rice to those in need during the pandemic-induced citywide lockdowns.xiv These rice ATMs appeared in Ho Chi Minh City, Da Nang, and other cities as part of community efforts led by universities and local Red Cross societies. The Ho Chi Minh City-based inventor of the rice ATMs went on to replicate face mask dispensing machines a few months later as COVID-19 cases surged in a second wave. The machines provided each recipient with three face masks, which could be washed and reused at least 30 times.xv
To prevent overcrowding, minimize surface contact, and verify identity, these dispensers relied on facial recognition technology linked to personal information, based on prior registration. Despite the risks to privacy, confidentiality, and other use of personal data, this technology did not prompt much public or government concern at all. This could be explained by the more pressing considerations of a public health crisis, compounded by food security uncertainties. With increased public awareness, there may be more questions posed about the longer-term implications of facial or other biometric recognition technologies, as well as the need for much more transparency, explainability, and accountability.
Interestingly, labor displacement as a result of an increased deployment of AI was identified as an ethical concern during stakeholder consultations. Survey data from Viet Nam’s 2019 population and housing census indicates that 80.8 percent of Vietnamese aged 15 years and above possess no technical or professional qualifications. Only 23.1 percent of the labor force does.xvi Studies have found that given the country’s prevailing labor structure, the impact of AI on individual workers could be challenging for a comprehensive, nationwide digital transformation without accompanying initiatives to reskill, upskill, or build a higher-quality labor market in line with at least regional, if not international, standards.xvii
There will also need to be special attention paid to the gender dimension of labor-displacing AI technologies. A gender-disaggregated analysis in 2021 showed that female workers tend to be slightly more affected than their male counterparts, in line with other research findings on the general impact of digital technologies on women, especially in developing economies.xviii
One informant pointed out that apart from the risks of unemployment and a consequently widening inequality gap, AI-supported innovations such as self-driving cars could disrupt the country’s budding gig economy—characterized by short-term, flexible, or freelance jobs such as ride-sharing or food delivery drivers—particularly in the aftermath of the COVID-19 pandemic.xix A 2021 survey of more than 60,000 workers in Viet Nam found that up to 53 percent of knowledge workers participated in the gig economy and only 40 percent of respondents wanted to return to the office.xx
In order for Viet Nam’s AI-driven digital transformation to be as competitive and inclusive as possible, it will need to overcome language and gender barriers. The realities of an English-dominated global science and technology field mean that if Viet Nam’s foray into the Fourth Industrial Revolution is to attract greater foreign investment, promote international cooperation, and effectively project the country’s own potential onto the regional and world stages in line with the Strategy, the nation will have to speak the common language of technology and business.
In 2008, the government launched Project 2020 with an approved budget of VND9.4 trillion (approximately USD443 million). The project’s vision was for most Vietnamese school, college and university graduates to be able “to use a foreign language confidently…in an integrated, multi-cultural, and multi-lingual environment, making foreign languages a comparative advantage of development for Vietnamese people.” However, in 2016, four years shy of its target date, the Minister of Education and Training, Phùng Xuân Nhạ, declared the project a failure before the country’s National Assembly.
Since AI material is primarily taught in English, language-learning support will be crucial not only for general communication purposes but also for a more technically-focused immersion into AI subjects. It would also help facilitate translation or interpretation of English technical terms such as “algorithmic bias” into everyday Vietnamese for widespread comprehension and more meaningful domestic conversations about the potential benefits and harms of AI.
Part of reducing the harms—particularly the risks of exclusion, bias, and discrimination—of AI technologies is to ensure that different segments of the country’s population are adequately and securely accounted for in building the datasets that train AI algorithms. Although female representation in Viet Nam’s National Assembly reached 26.7 percent in the course of its 2016-2021 term and women held 12 of 30 ministerial and deputy ministerial positions in 2017,xxi Vietnamese women still fall short of adequate representation at the C-suite level in the business world. In 2020, one survey found that women comprised only 17 percent of board members, 12 percent of chairs, and nine percent of chief executive officers in all Vietnamese listed entities.xxii At the university level, women account for only 36.5 percent of science, technology, engineering, and mathematics (STEM) tertiary graduates compared to 63.5 percent of men.xxiii The good news is that the margin of difference between women and men researchers currently in STEM is slightly lower at 44 percent compared to 56 percent.xxiv
The gender imbalance in STEM fields and in corporate boardrooms translates into the very real exclusion of half the country’s population, along with their perspectives and realities in the design and development of AI technologies as well as the decision-making processes of society’s digital transformation. It also sits awkwardly with the fact that most of Viet Nam’s enterprise landscape is dominated by women-owned SMEs and female consumers.xxv
Rather than an afterthought, ethics should be baked into the design of AI algorithms and systems from the very outset. This means identifying the principles, values, and objectives of AI use in Viet Nam; being familiar with the terminologies of ethics in the global discourse; adapting that language to the domestic context, as appropriate; teaching AI ethics in engineering and other related courses at universities; and incorporating a cross-disciplinary approach to AI technology. Indeed, the Strategy tasks the Ministry of Science and Technology with promoting the formation of “open professional groups for multi-, inter-, and trans-disciplinary use and exchange of AI data technologies.”
Additionally, the Ministry of Information and Communications will develop legal and policy frameworks to ease data sharing and AI testing in sandboxes. Importantly, it will “develop standards, technical regulations, and formats for AI technologies and products.” Yet, Viet Nam is currently neither a participating nor observing member of the ISO and IEC Joint Technical Committee on Artificial Intelligence, which has published 11 ISO standards and is developing 26 more in the area of AI, including in the treatment of bias as well as related ethical and societal concerns.
Rather than adopt an existing model wholesale, it appears that Viet Nam may well chart its own AI and ethics path to suit its local setting. Although the OECD Principles on Artificial Intelligence and accompanying recommendations were forwarded to the Vietnamese prime minister’s office for consideration,xxvi a government release in 2019 soberly noted that, “if Viet Nam follows the trend of AI development like developed countries with solid AI resources, it will be difficult for Viet Nam to keep up with them…. Therefore, Viet Nam must have its own direction for the AI industry.”xxvii
Our informants summarized it best: Viet Nam will have to take a step-by-step approach to integrating ethics in AI application. That process begins with an increased awareness of the topic and a greater exchange of perspectives domestically and regionally. The country’s ambition of being a developed economy warrants reflection on the standards already proposed and embraced by more technologically mature states. Nevertheless, Viet Nam will also study alternative models, and look to cooperate with its ASEAN neighbors to adapt the existing suite of AI principles for an optimal domestic and regional fit. For now, while there is some degree of recognition in Viet Nam of the importance of AI ethics, their formulation and application remains underexplored.
i Soon Ghee Chua and Nikolai Dobberstein, “Racing Toward the Future: Artificial Intelligence in Southeast Asia,” Kearney, accessed June 1, 2022, https://www.kearney.com/digital/article/?/a/racing-toward-the-future-ar…;
ii “Decision No. 127/QD-TTg 2021 National Strategy for Artificial Intelligence Research through 2030,” LuatVietnam, January 26, 2021 https://english.luatvietnam.vn/decision-no-127-qd-ttg-dated-january-26-….
iii Phuong Nguyen, “’Make in Vietnam’ campaign targets top 30 IT status,” VNExpress, May 8, 2019, https://e.vnexpress.net/news/business/industries/make-in-vietnam-campai…;
iv “Make in Vietnam, by Vietnam for a fresh digital orientation,” Vietnam Investment Review, December 31, 2020, https://vir.com.vn/make-in-vietnam-by-vietnam-for-a-fresh-digital-orien…;
v “Viettel AI Open Platform: Shaping a digital society on artificial intelligence platform,” Viettel, accessed June 1, 2022, https://viettelgroup.ai/en.
vi “Zalo Debuts First Vietnamese AI Assistant Named Kiki,” Saigoneer Soceity, January 19, 2021, https://saigoneer.com/saigon-technology/19872-zalo-debuts-first-vietnam…;
vii “Change the Language of Google Assistant,” Google Nest Help, Accessed June 1, 2022, https://support.google.com/googlenest/answer/7550584?hl=en&co=GENIE.Pla…;
viii In a 2020 survey of 10,197 enterprises carried out by the Viet Nam Chamber of Commerce and Industry, 87.2 percent responded they had suffered "mostly" or "completely negative" impacts from the pandemic. In terms of labor redundancy, 35 percent of private companies and 22 percent of FDI companies reported reducing their workforce. Among the former, micro- and small enterprises laid off the most workers (36 percent and 35 percent, respectively). https://vcci-hcm.org.vn/wp-content/uploads/2021/03/WB-VCCI_Covid-19-Rep…;
ix “Make-in-Vietnam key to making VN a digital powerhouse,” Vietnam net Global, January 20, 2022, https://vietnamnet.vn/en/make-in-vietnam-key-to-making-vn-a-digital-pow….
xi Cameron A, Pham T H, Atherton J, Nguyen D H, Nguyen T P, Tran S T, Nguyen T N, Trinh H Y and Hajkowicz S, “Vietnam’s Future Digital Economy Towards 2030 and 2045,” CSIRO Research, January 12, 2022, https://research.csiro.au/aus4innovation/wp-content/uploads/sites/294/2…; Nga Than and Khoa Lam, “Challenges of AI Development in Vietnam: Funding, Talent, Ethics,” MAIEI, September 21, 2021, https://montrealethics.ai/challenges-of-ai-development-in-vietnam-fundi…;
xii Stakeholder consultation; “Background on LAWS in the CCW,” United Nations, https://www.un.org/disarmament/the-convention-on-certain-conventional-w…;
xiii See, e.g., Will Douglas Heaven, “Predictive Policing Algorithms are Racist. They Need to be Dismantled,” MIT Technology Review, July 17, 2020, https://www.technologyreview.com/2020/07/17/1005396/predictive-policing…; Odhran James McCarthy, “AI & Global Governance: Turning the Tide on Crime with Predictive Policing,” United Nations University Center for Policy Research, February 26, 2019, https://cpr.unu.edu/publications/articles/ai-global-governance-turning-….
xiv “Free ‘rice ATM’ for those in deed in Da Nang amid Covid-19 crisis,” Da Nang Today, August 19, 2020, https://baodanang.vn/english/photo-news/202008/free-rice-atm-for-those-…;
xv Josee Ng, “Saigon Man Invents Free Mask ATM to Help Residents Protect Themselves from the Coronavirus,” The Smart Local Vietnam, August 7, 2020, https://thesmartlocal.com/vietnam/face-mask-atm/.
xvi “20 Major Indicators of the 2019 Population and Housing Census,” General Statistics Office, December 19, 2019, https://www.gso.gov.vn/en/data-and-statistics/2019/12/infographic-20-ma…;
xvii Nguyen Minh Tru and Doan Thi Nhe, “Impact of Industrial Revolution 4.0 on the Labor Market in Vietnam,” Research in World Economy 12, no. 1 (2021): 94-100, doi: 10.5430/rwe.v12n1p94
xviii Francesco Carbonero, Jeremy Davies, Ekkehard Ernst, Frank M. Fossen, Daniel Samaan, and Alina Sorgner, “The Impact of Artificial Intelligence on the Labour Market in Lao People’s Democratic Republic and Viet Nam,” United Nations: 1-36, August 9, 2021. ; Judith Mariscal, Gloria Mayne, Urvashi Aneja, and Aline Sorgner, “Bridging the Gender Digital Gap,” Economics Discussion Papers, no. 2018-60 (August 28, 2018), Kiel Institute for the World Economy.
xix Stakeholder consultation.
xx “Only 40% of Vietnamese workers willing to return to office: survey,” Tuoi Tre News, December 23, 2021, https://tuoitrenews.vn/news/business/20211223/only-40-of-vietnamese-wor…;
xxi “Viet Nam’s Voluntary National Review on the Implementation of the Sustainable Development Goals,” United Nations, June 2018, https://sustainabledevelopment.un.org/content/documents/19967VNR_of_Vie…, p. 11. For more on Viet Nam’s gender equality profile, see “Country Gender Equality Profile Viet Nam 2021,” United Nations Entity for Gender Equality and Empowerment of Women, 2021, https://www.ilo.org/wcmsp5/groups/public/---asia/---ro-bangkok/---ilo-h…;
xxii “Progress at a Snail’s Pace: Women in the Boardroom: A Global Perspective,” Deloitte Global Boardroom Program, 2022, https://www2.deloitte.com/content/dam/Deloitte/global/Documents/gx-wome…;
xxiii “Gender Statistics,” The World Bank Data Bank, accessed June 1, 2022, https://databank.worldbank.org/id/2ddc971b?Code=SE.TER.GRAD.FE.SI.ZS&re…;
xxiv “A Complex Formula: Girls and Women in Science, Technology, Engineering and Mathematics in Asia,” UNESCO Bangkok, May 4, 2017, https://bangkok.unesco.org/content/complex-formula-girls-and-women-scie…;
xxv “VN Needs More Women Working in STEM: experts,” Viet Nam News, November 14, 2017, https://Viet Namnews.vn/society/417457/vn-needs-more-women-working-in-stem-experts.html.
xxvi Stakeholder consultation.
xxvii “Selecting Appropriate Artificial Intelligence Development Strategy,” Ministry of Information and Communications of the Socialist Republic of Vietnam, August 28, 2019, https://english.mic.gov.vn/Pages/TinTuc/139578/Selecting-appropriate-ar…;