Human capital futures in the global south: Employability and education policy in the age of AI

Introduction

Human capital has its conceptual origins in the Global North through the writings of the British economic philosopher Adam Smith in the 1700s and later developed by American economists Theodore Schultz, Gary Becker, and Irving Fisher throughout the 1960s to 1980s (Le Chapelain, 2024). Given its starting point in the Global North, I contend that the utility of human capital within the Global South suggests a reframing of its emphasis. This is partly due to differences between the Global North and Global South that are evident historically, materially, socially, and otherwise. Legacies of inequality and exploitation often define the fault lines between nations, which constitute the Global North–South divide (Kamal Pasha et al., 2024).

In this article, I argue for a reframing of the concept of human capital in Global South contexts such as the Association of Southeast Asian Nations (ASEAN) ¹ , specifically through higher education policy. I emphasize the Vietnamese context as a key example by highlighting State interests concerning human capital formation. This article raises the question: in what ways Vietnam’s higher education policy discourse constructs a view of human capital formation compatible with the age of AI? My discussion engages with the complications of the recent development of generative artificial intelligence (AI) relating to higher education policy advances.

The promise of AI in global education is dominant in global education discourse while the perils of AI remain marginal (Lin, 2023; Rahiman and Kodikal, 2024). As such, I problematize technologically centered education development which can potentially undermine higher education policy and productive education-related outcomes in the Global South. This article contributes to the field of critical policy studies, emphasizing education in Global South settings. I maintain that robust Global South human capital development through education needs to remain human-centric, yet responsive to ongoing transformations concerning AI technologies.

Education policy is defined as a discursive apparatus of educational leadership that specifies principles, actions, and outcomes concerning education-based purposes (Trowler, 2003). Lee (2016) observes that human capital is often produced through educational institutions, being both “scarce and valuable” in terms of how this positions local institutions and countries globally (p. 466). For Global South universities, education policy serves as a framework for ensuring education meets the unique needs of their populations. Higher education is integral to human capital development as it elevates the standard and quality of socioeconomic outcomes in modern knowledge-based societies (Liu et al., 2024).

Higher education policy is important in establishing structure, priority, and directives for the ways this level of education contributes to human capital development. According to Bell and Stevenson (2006) “educational leadership does not exist in a vacuum—it is exercised in a policy context, shaped decisively by its historical and cultural location” (p. 7). This understanding further solidifies the contextual production of human capital formation through higher education. Differences in context account for differences in employability and the kind of education needed across Global South territories such as ASEAN and countries such as Vietnam in particular.

Human capital, AI and higher education

Human capital represents a conceptual framework concerning the relationship between one’s education and labor productivity. This concerns a range of education-based qualities and competencies a person develops throughout their formal schooling (Flabbi and Gatti, 2018). Furthermore, human capital forms part of a wider range of capitals concerning the notion of employability, which also constitutes social, cultural, identity, and psychological capitals (L. H. N. Tran et al., 2020). As such, human capital serves as a metric for socioeconomic progress, being a mainstay of 21st-century discourses on employability (Nghia et al., 2020; Schueller, 2023). Scholars observe “an emerging interest in how human capital can be reconceptualised” given changes across global higher education and employment landscapes (Donald et al., 2024, p. 10).

Studies show that employability concerns not only one’s knowledge, education, and skills, but also the ways this is communicated by individuals navigating education and career paths (Gorbatov et al., 2024). Contemporary human capital has also been identified by experts as being multifactorial and nonlinear in its development (Sulaiman et al., 2024). These contemporary conceptions of human capital intersect with developments in higher education which plays a vital role in sustainable futures. However, I add that AI further complicates the relationship between higher education and human capital development. This complication concerns the nature of this disruptive technology, perceived global value, and how it has and continues to upend higher education worldwide.

Generative AI can be defined as digitally enhanced systems capable of producing various kinds of content such as text, still, and moving image, as well as audio using human input (Feuerriegel et al., 2024). In late 2022, the large language model application ChatGPT became an international sensation being perhaps the most sophisticated and accessible text-based generative AI system known at that time (Hu, 2023; Tlili et al., 2023). Since then, the emergence of similar AI platforms has followed, including Deepseek, Grok, Gemini, and Bard, allowing users the opportunity to efficiently generate outcomes with little to no cognitive computation or creative process (Jegham et al., 2025; Leong et al., 2025). Generative AI often draws from online datasets and digitized repositories of information produced by human agents. Such systems aim to replicate and reconfigure various units of information to “creatively” produce or mimic work of the human agents (Bandi et al., 2023). On the one hand, AI holds potential for enhancing human capability and by extension employability. Yet I maintain that there is greater danger for AI to undermine human capital formation as it often shortcuts processes in favor of outcomes.

While the potential of AI is a tantalizing prospect, the contemporary discourse on AI decenters human capability in favor of neoliberal efficiency (McInnes, 2025). Kalluri (2020) argues against the neutrality of AI as these technologies as they “sanctify the status quo and advance the interests of the powerful” (169). Acemoglu and Johnson (2023) note that AI superfluously replaces human labor often as a means of reducing labor costs or increasing production efficiencies. The commercial nature of AI technologies and their superfluous potential in mimicking human activity positions it as an exploitive and extractive invention. Malik (2025) argues that AI works as a neoliberal replacement of human competence rather than an aid to it considering the amplified commodification of knowledge in modern universities. It is undeniable that the most dominant AI technologies originate from the Global North (Yu et al., 2023). The social, economic, and political benefits of AI remain concentrated in this region also, further complicating Global North–South divide. Global human capital formation is impacted by the impact of AI on education as disruption has not always proved positive.

There is growing evidence of AI in education negatively affecting issues of intellectual integrity, literacy, and critical thinking (Cassidy, 2024; Hayes et al., 2024; Weale, 2025). A study by Bai et al. (2023) highlighted both positive and negative effects of AI in education yet raised significant concerns about “the potential deterioration of critical thinking skills” (p. 8). Research by Jose et al. (2025) identifies AI in education as “both a cognitive amplifier and inhibitor” that can accelerate learning progress and regression (p. 3). Gerlich (2025) sees the benefits of AI in education in terms of “efficiency and accessibility” yet signals caution in its ability to severely impair individual capacity and by extension human capital development. Collectively, Bai et al. (2023), Gerlich (2025), and Jose et al. (2025) all emphasize the role of policy in mitigating the negative effects of AI in education as this is inevitably connected with human capital formation.

Chiocco and Farrell (2025) argue against a pro-AI approach to education as “lowering standards of rigor for everyone helps no one” (para. 11). Research by Corbin et al. (2025) highlights the need to restructure higher education to demonstrate students’ competence considering the challenges of AI regulation. Research has also confirmed a negative impact on academic outcomes for higher education students with the use of generative AI technologies such as ChatGPT (Palmer, 2024). Taylor (2024) observes that AI potentially “undermine(s) the trust needed between students and instructors” (para. 5). Other scholars also identify academic dishonesty as a major concern regarding generative AI technologies in higher education (Cotton et al., 2023). Despite the potential of AI to positively transform education worldwide, there is growing concern about its negative repercussions which seemingly far outweigh its benefits. This is further amplified given the Global North–South divide which further threatens employability prospects in developing countries.

Given these legitimate concerns, I add that there is also the concern of underdeveloped competencies in university students due to an overreliance or unethical use of AI in cases of plagiarism. This in turn undermines the potential of human capital development via higher education. AI is conspicuous in that it draws attention to its capabilities rather than what human agents can do. Despite the efficiencies that AI affords, it is through social processes that human capabilities are developed, not through shortcuts to producing outcomes. In the global quest for human capital development, AI presents a mode of transport to a destination without a necessary journey where ends can be achieved without due consideration for means. AI is now an unavoidable consideration in worldwide higher education in the 21st century (Makeleni et al., 2023; Rudolph et al., 2023). Both strategic and tactical tools are needed to enable the production of high-quality human capital through higher education in these settings.

Global South studies confirm that productive AI in these settings requires a “rearrangement of their education” to experience the betterment these kinds of technologies can provide (Heng et al., 2022, p. 2). Just as scholars previously critiqued the internet as, by default, an emancipatory tool of Global South advancement (Curran et al., 2012), Prinsloo (2020) also argues that emerging technologies of big data can and often do reinforce a problematic status quo, citing the political economy of the internet and the hegemony of the Global North concerning digital ecosystems. Another study on ChatGPT concludes that “AI technology may seriously widen the education gap like no other technology before it” (Kasneci et al., 2023, p. 8).

Blind faith in AI does little to address education gaps and is likely to exacerbate them. Improving education for human capital futures for the Global South in the age of AI requires both bottom-up and top-down solutions. In terms of the latter, experts worldwide contend that education policy is important to 21st century sustainability, security, and survival (Carstensen and Emmenegger, 2023; Custers and Magalhães, 2021; Olssen et al., 2004). This is especially true of Global South countries which are faced with limitations and threats to these aspirations which are markedly different to the Global North. Global South governments understandably aim to leverage the prospects of AI in their education systems.

The unstable nature of AI complicates straightforward ways of addressing Global South challenges. This makes future-oriented, technologically driven education policy dubious if an overdependence of AI is central to it. The potential of AI does not automatically provide reprieve for existing challenges within Global South higher education, as these issues operate at the “interplay of law, culture, organizational forms, social norms, legitimacy and functionality of institutions” (Vögtle and Windzio, 2025, p. 904). ASEAN countries largely exemplify common challenges in the Global South. Such challenges include issues of intra-ASEAN territorial conflict, civil unrest, population aging, cybercrime, and human trafficking among others (Acharya, 2013; Sundram, 2024). The populations of these countries are ethnically and linguistically varied in their constitution and the challenges they face. Higher education plays a key role in mitigating many of these challenges and providing human capital to address them directly.

ASEAN, Vietnam, and education policy

Studies on human capital in ASEAN have observed the importance of curriculum and education quality as means for sustainable futures for countries in the region (Goujon and Samir, 2008). This is an important consideration given that higher education is highly contextual. Education is key to the development of these nations’ human capital, with several studies emphasising the role of human capital in ASEAN’s future (Haini, 2019; Hong Vo et al., 2021). Scholars have also observed the need for a more contextualized understanding of human capital in the Global South, looking at ASEAN countries in particular (Crocco and Tkachenko, 2022; Houston, 2024). Infrastructure and technologies are also factors which impact human capital formation in a wider sense, as higher education policy needs to consider the environment in which human capital is developed. Vietnam serves as one of the more recent members of ASEAN having joined in July 1995 (Narine, 2008).

Like other countries in the Southeast Asian bloc, Vietnam has aligned its higher education system according to the shared quality assurance standards globally, including the ASEAN University Network-Quality Assurance established in 1998 (H. C. Nguyen and Ta, 2018; Pham and Nguyen, 2023). Since the early 2000s, the Vietnamese government has emphasized a general internationalization of national policy across a range of fields, with higher education included (Hoai et al., 2023). This trajectory led to the development of several key policy instruments including the general Education Law (Education Law, 2005), Higher Education Reform Agenda (HERA) of the same year (Hai et al., 2023), Higher Education Law (2012), and its subsequent amendment (2018). The work of Tran and Tran (2021) provides a detailed account of changes within the Vietnamese education policy discourse from 2005 to 2019. This sheds light on the opportunities and challenges faced with the development of Vietnam’s higher education system.

Despite these policy advancements, research indicates that higher education in Vietnam still inadequately prepares its students for industry, international or regional exchanges thereby limiting its human capital potential and global competitiveness (S. L. Bengtsson, 2022; Hill et al., 2021; L. T. M. Nguyen et al., 2021). Vietnam has notable issues with its human capital formation unlike many of its ASEAN neighbors. For instance, Vietnam’s industrialization rate has been 15–20% lower than other ASEAN countries (Fukuoka, 2021). Also, the nation has persisted with low labor productivity compared to other countries in Southeast Asia, which in partly related to education quality (Q. T. T. Le et al., 2020; D. Phan, 2024). Considering these matters, AI has also become a more visible part of Vietnam’s higher education landscape. However, this does not automatically translate to better education outcomes.

Vietnamese higher education and AI

There exists an emerging discourse on Vietnamese higher education and AI concerning its utility and benefit in a wider range of fields. In 2019, the local press reported the government’s aim to make AI a top policy priority (Viet Nam News, 2019). This included the development of local human capital including the corresponding improvement of higher education. Following this, a State policy concerning digital transformation was approved in 2020, which concerned the use of AI for priority areas such as banking, transport, and health (Mai, 2023). Collaborations with educational institutions were seen as key to capitalizing on the potential of AI technologies. Rivalling the popularity of the online text-based ChatGPT from the Global North, Vietnamese tech company ETT International JSC engineered a competing voice-based robotic system called Anan (Trong, 2023).

The AI ambitions of the Vietnamese government caught the attention of global companies such as Ericsson, which pledged to invest in a local AI education laboratory in an international university within the nation’s capital of Hanoi (Donkin, 2023). Similarly, in 2024, authorities in the southern economic hub of Ho Chi Minh City have planned for the closer integration of AI technologies in local education. In late 2025, the State approved its first ever Law on Artificial Intelligence which took effect in March 2026 (Dung, 2025), however this legislation is not centered on higher education. However, Vietnam’s race for AI development in education and other sectors has been met with caution by some. Vietnamese university leaders and other members of the local business community have pointed out that AI can undermine the nation’s human capital (Van, 2023). I add that this sense of caution is warranted, given what generative AI constitutes and existing studies concerning its use in Vietnamese education.

Local experts such as Quy et al. (2023) see AI in Vietnamese higher education as part of a greater national effort towards digital transformation. However, such developments are not uniformly positive. Duong et al. (2023) examine the potential of ChatGPT concerning student intention and perceived personal benefit. In this study, while the benefits of AI in higher education are not explicitly stated, they do emphasize appropriate policy-based responses to “the potential of AI-driven tools like ChatGPT” (Duong et al., 2023, p. 375). Maheshwari (2023) reiterates the need for the “appropriate and ethical use of AI tools” in Vietnamese higher education, considering matters of assessments which gauge students’ competencies (p. 12189).

Another study on ChatGPT use among Vietnamese university students raised concerns regarding “academic dishonesty, as well as reduced creativity and critical thinking” (Nguyen et al., 2024, p. 116). In this case, students were unsurprisingly more likely to engage in cheating which impeded their academic development. Similarly, H. M. Nguyen and Goto (2024) highlighted the need for “(e)ducational policies for promoting academic integrity” given the prevalence of ChatGPT among higher education students (p. 17). They also highlight the need for a more “stringent” policy framework to discourage “AI-powered academic cheating” (Nguyen and Goto, 2024, p. 17). Evidently, the emerging discourse on Vietnamese higher education and AI strongly suggests the importance of policy in constructing desired realities on human capital development. Considering this understanding, I now discuss the theoretical framing of this study including how Global South higher education policy can maintain robust human capital formation given the practical and ethical complications of generative AI.

Theory

In this study, I adopt a poststructuralist stance, which is a theoretical viewpoint that considers the conditional nature of knowledge and practice, predominantly shaped by social and historical factors. Given (2008) explains that poststructuralism emphasizes “identifying meanings that are context-specific and that relate to the varying discursive practices” (p. 2). While poststructuralism is not a unified system of thought or a collective body of theory, Williams (2005) acknowledges that poststructuralism is characterized by the fact that “the limits of knowledge play an unavoidable role at its core” (p. 1). Dillet et al. (2013) also observe that poststructuralist thought rejects universality and fixture, instead emphasizing “the limits of existing knowledge, truths and values” (p. 5).

Poststructuralism problematizes the idea of “certainty” not by through arbitrary or subjective relativism but by providing an account of difference and change as it occurs and is experienced across time and space. As such poststructuralism is resistant to essentialism, while not denying the structuration of the material and social worlds humans navigate. This theoretical perspective is a useful lens considering the ways through which AI advances remain in flux and how Global South education policy can productively respond changes in human capital formation.

Concerning the application of poststructuralism to policy analysis Bacchi and Goodwin (2016) state that this paradigm aims to address a plurality of realities that are “contingent, open to challenge and change” (p. 4). As such, my application of poststructuralist thought in education policy analysis considers how challenges and opportunities are constructed and addressed. Challenges and opportunities are not naturalized or inherent; rather they are defined and understood based on where these might emerge at any sociohistorical moment. With this understanding of poststructuralism, I use this theoretical framing in further de-essentializing what human capital constitutes, especially in the Global South, and the implications for higher education policy in this setting. This de-essentializing especially considers the unavoidable problematic of generative AI, as this technology marks a new historical period globally with notable implications for how higher education is conceptualized (Aoun, 2017; Hayes et al., 2024; Jafari and Keykha, 2024).

The texts I examine in this study are neither “neutral,” nor embody a particular “objective truth” regarding human capital formation through higher education. Given the context and purpose of the policy instruments analyzed in this article, I argue that human capital development via Vietnamese higher education policy presents productive considerations for student employability in the age of AI. Such policy focus elevates human agency over deterministic technologically centered education development which I have argued against elsewhere (Felix, 2021). The policy instrument featured in this study is representative of sociopolitical interests of the State which prioritizes a human-centered approach to its local human capital formation. My analysis of Vietnamese higher education policy spotlights the importance of de-centering education policy from an unstable technological core.

Method

I employ Critical Policy Discourse Analysis (CPDA) to examine Vietnam’s Higher Education Law (2012) and its later amendment (2018). I present my findings highlighting select portions of this State policy as it centers on sustainable, human-centered, human capital development through higher education. CPDA is a method entailing a systematic investigation of policy instruments looking at matters of structure, language, and signification and how these construct the empirical realities they represent (Mulderrig, 2024). Policy instruments here concern the ways higher education is institutionally regulated through officially published statements of intent and implementation. Policy can work at the level of the State or the level of individual education institutions. Scholars use the term “policy instrument” to designate “what governments do” and “the core businesses of public policy analysts” (De Boer and Van Vught, 2020, p. 1). Similarly, Ali (2013) argues that policy instruments are “governing tools…to achieve policy targets” (p. 99). I employ the term policy instrument as an analytic category, considering the ways these texts function including the ways these construct the realities of people, problems, and contexts they are meant to represent (M. Bengtsson et al., 2010).

CPDA also concerns issues of power that operate through policy. In this regard, policy contributes to and constitutes the construction of discourses which enable and sustain relations of power among various entities and actors in a given environment. Stuart Hall’s (2019) explanation of discourses is useful here as it articulates the basis of the textual methodology which underpins this study:

Discourses are ways of talking, thinking, or representing a particular subject or topic. They produce meaningful knowledge about that subject. This knowledge influences social practices, and so has real consequences and effects. Discourses are not reducible to class interests but always operate in relation to power— they are part of the way power circulates and is contested (p. 160).

As such, discourses refer to the conceptual structuring of the empirical world using language and other modalities of communication. Issues of power relate to the origin of policy and the ways it enables or constrains the agency of human subjects and social entities. This also includes the parameters, challenges, and opportunities, notwithstanding how these are delineated and reinforced through policy. CPDA is derived from the established method of Critical Discourse Analysis and the field of Critical Policy Studies (Mulderrig et al., 2019), yet maintains emphasis on policy instruments. Concerning this textual method Papadopoulos & O’Keeffe (2024) assert that the analysis of policy is premised on the idea that “social processes and discourses are mutually constitutive” and that “discursive framing” accompanies the social changes which is a result of policy.

In the context of this study, the Vietnamese Higher Education Law (2012) and its later amendment (2018) constitute a localized discourse concerning its human capital development through higher education with ramifications for graduate employability in the age of AI. The analysis presented in this research entails a 3-step process of analysis which includes detailing the context and features of this policy instrument, then supplying a reading of it in relation to AI and employability and finally offering an evaluation of the implications of such.

Analysis

Vietnam’s official higher education policy was produced before the mainstreaming of AI in 2022 and its inescapable conspicuity in the global education landscape. This law is also distinct from its AI law enacted in 2026, which aims to regulate general AI usage and development locally. Vietnam’s Higher Education Law of 2012 is the first significant piece of legislation precisely directed towards matters of higher education, with its purpose to regulate the development of related institutions in the process of teaching, research, and subsequent human capital formation. However, the historical trajectory of this policy instrument is decentered from technology as a goal or tool. Instead, it is a means that fosters greater human agency among the Vietnamese citizenry.

In 2018, the first and only amendment to the 2012 Higher Education Law sought to meet “the demands of globalization, industrialization, and modernization” (Nguyen and Tran, 2019, p. 184). It addressed undeveloped areas including governance structures for university autonomy and tuition financing, research funding for institutions regardless of classifications or rankings, and consistency of the application of the VQF (C. H. Nguyen and Shah, 2019; L. H. Phan and Doan, 2020). This legislative amendment was positive in moving Vietnam’s higher education sector towards a more “modern governance” structure (Parajuli et al., 2020, p. 3).

Highlights of this legislation include the classification and ranking of universities, university autonomy, university research capacity, quality assurance, and internationalization (Law on Higher Education, 2012; On Amendments to the Law on Higher Education, 2018). These elements are related to the development of local human capital which constitutes graduate employability. No other significant piece of legislation exists that is exclusive to higher education. Vietnam’s Higher Education Law was an effort to create a “multi-tiered greater education sector with top-tier research-intensive universities” (Phùng, 2020, p. 227). Universities were to be first ranked locally based on prestige, in keeping with their quality mechanisms to foster employability (L. H. Phan and Doan, 2020). This first instance of the Education Law consolidated an assortment of previous regulations since the early 1990s. It moved universities away from being inward-focused, monodisciplinary entities towards being diversified and internationalized (C. H. Nguyen and Shah, 2019; N. T. Nguyen and Tran, 2019).

Such regulations were aimed at shaping empirical realities of employability via a conceptual starting point through policy. In this instance, local and global employability became a notable consideration in a shift away from a marked focus on disciplinarity. Considering these policy advances, technology was never central to Vietnam being modern, internationalized, globalized, or industrialized. Despite the rapid changes brought about by AI technologies, Vietnam’s education policy subjects technology to being a means to an end rather than an end in itself. This policy exemplifies the importance of human-centered education policy rather than one led by unstable technological development underpinned by neoliberal efficiency. Vietnam presents a case of Global South higher education policy that does not have a technocentric or technologically deterministic character.

The wider context of the Higher Education Law and its amendment was concurrent with broader State initiatives for social and economic development that emphasized Vietnam being a “modern, industrial country” in addition to “building a popular, nationalistic, advanced, modern, socialist education” in its orientation (Tran and Marginson, 2018, p. 2). Through this legislation, students were to be modeled good practice from faculty, concerning self-improvement through political knowledge, academic specialization, pedagogical expertise, and Confucian values. This constituted the “material and spiritual conditions for teachers to fulfill their roles and responsibilities, preserving and developing the tradition of respecting teachers and honoring the teaching profession” (Le and Hayden, 2017, p. 84). This orientation is human-centered, situating human agents as the primary drivers of change for the realization of a modern and creative Vietnamese society. Modernity in Vietnam’s education policy is not constrained to a deterministic vision of technocentrality, but subjects this to a people-centered value system rooted in tradition and heritage. While AI may have valuable uses, this is outweighed by its unstable nature, commercial underpinnings, and instrumental optimism often ascribed to it.

Education policy scholars have highlighted the significance of repetition in their readings of policy instruments (Hayes, 2019; Serrano-Velarde, 2015). The use of this type of rhetoric reinforces the validity of the use of these terms, normalizing their inclusion in this policy instrument and the priority of the policy actors these terms come to represent. In this regard, the use of the term “technology” occurs 56 times in the original version of the Vietnamese Higher Education Law (2012) and 14 times in its later amendment (2018). However, considering the number of times this term is referenced, Vietnam’s higher education policy is not technology-centered nor is it technologically oriented.

Most instances of the term “technology” refer to its use in a disciplinary sense as a subject of pedagogy or as an emphasis of a particular faculty. In other instances, the term is closely linked with that of “science” in generally the same sense. Examples of this can be found throughout this policy instrument as presented here (emphasis mine):

This Law specifies the organization, duties and authority of higher education institutions, the activities of training, science and technology, international cooperation, higher education quality assessment and assurance, the lecturers, the students, the higher education institutions property and finance, and the State management of higher education (Socialist Republic of Vietnam, 2012b, p. 1).

In contrast, several terms referencing the idea of human capital are actively used to frame the contents of this legislation by identifying the aim of Vietnamese higher education and those whom it is meant to serve. These references to the nation’s human capital development are presented at macro, meso, and micro levels. These references are often stated via the terms such as “human resources,” “socioeconomic development,” and “competitive capacity” in a macro sense. Similarly, at meso and micro levels, the concept of human capital is evident through students’ “capability of researching” and “professional responsibility and adaptability,” which includes “employer’s requirement.” Examples of this can be found in the introductory section of this policy instrument and in key sections which reemphasize its purpose (emphasis mine):

Training human resources, enhancing people’s intelligence; doing science and technology researches [sic] in order to create knowledge and new products serving the socio-economic development, assure National defense and security and international integration (Socialist Republic of Vietnam, 2012b, p. 2).

Moreover, Vietnamese higher education policy humanizes its approach to education via 3 key areas: workforce development, social mobility, and citizenship. In previous work, I have argued for the significance of these key areas of Vietnamese human capital formation through its higher education system (Felix, 2024). Local higher education policy, in its present state, places greater emphasis on people than on technologies. The idea of workforce development is present in the case of explicit references to “satisfying the workforce demand” among similar terms and ideas expressed throughout (Law on Higher Education, 2012, p. 4). Social mobility emphasizes improving the life quality of citizens, especially rural ethnic minorities, for example, “key economic regions and impoverished areas” throughout (Socialist Republic of Vietnam, 2012b, p. 4). In addition, citizenship paying attention to the ethical, civic and intellectual dimensions of Vietnamese students e.g. participate in “social activities, environment protection, preserving the order and security” (Socialist Republic of Vietnam, 2012, p. 22). While optimism concerning the potential of AI to reshape education exists, the unstable nature of this technology does not make it a secure foundation for human capital development. Education policy need not ignore technological developments but ensure these are not amplified to supersede human considerations.

Issues of higher education are within the ambit of an established or recognized nation-State, with matters of citizenship being closely related to this also. Bayly (2014) argued that “Vietnamese educational reform initiatives are a critical arena for citizens’ experience” (p. 497), suggesting the difficulty of divorcing human capital from citizenship. Similarly, the employability of a nation’s populace is a workforce-related concern as this is also related to the quality of life experienced by citizens. While technology can be an aid to employability, the unstable and superfluous nature of AI redirects attention away from human capability and by extension employability. Generative AI is not a static technology, and the rapid pace of its development exacerbates its risks despite its benefits, as discussed earlier.

Notwithstanding Vietnam’s aims to leverage AI technologies to enhance its education, like other countries all over the world, its current higher education policy has made technology subservient to State interests rather than the center of its quest for sustainable development. The idea of sustainable development is especially prominent in the amendment to the Higher Education Law (2018). This change from the first iteration of the law signals closer alignment with “global” concerns and reflexivity concerning Vietnam’s status as a Global South nation. With the mission to holistically modernize its education sector, Vietnam’s initial higher education policy and its amendment maintains human-centeredness in positioning its university students as employable members of the national and global labor force.

Conclusions

This article aimed to address the question of how Vietnam's higher education policy discourse, as a Global South nation, constructs a view of human capital formation compatible with the age of AI. Responding to this question, I maintain that generative AI in higher education holds prospects for positive transformations, but even greater possibilities for undermining student employability. Highlighting the abstract potential of AI in closing educational gaps between Global North and South demands more attention be given to risks which outweigh such benefits. The negative potential of AI in education is to the detriment of human capital development across the globe, particularly in the Global South. Within such settings, limited mitigation exists for the risks posed by AI in addition to the Global North. In the case of Vietnam, its homegrown higher education policy was initially constructed with a human-centered approach to education. Employing Critical Policy Discourse Analysis (CPDA), in this study I evaluated the current orientation of Vietnam’s main higher education policy concerning its human capital futures.

This orientation is valuable considering that it maintains an emphasis on workforce development, social mobility, and citizenship. This is in keeping with the nation’s Confucian and socialist influences which place people at the center of education, with technology primarily framed as an area of disciplinarity. The Higher Education Law (2012) is still in effect to this date, along with select revisions through its amendment (2018). Key terms and phrases, expressed through repetition and/or through special placement within policy instruments, are indicative of Vietnam’s quest for a modern and creative society through its higher education reforms. While the State is keen on AI implementation within its higher education landscape, complementary policy instruments supporting this initiative should continue the trajectory of the current higher education policy.

In a broader sense, Global South higher education policy should be human-centered. It would be prudent for policymakers to resist the temptation towards adopting a deterministically positive outlook concerning AI technology. Such a perspective can be blinding to the need for the importance of information literacy and the ethics of technology which would enhance appropriate policy initiatives concerning AI in education. The popularity of AI technology does not negate its unrealized capacity to impair the employability of higher education students. While 21st century higher education policy need not turn a blind eye to the challenges and opportunities of AI, its problematic nature requires caution against an uncritical embrace of it.

Education quality in terms of human capital cannot exist without educational institutions which function optimally with appropriate policy as a roadmap for productive futures. In addition, the optimal functioning of education institutions occurs in large part by directives identified and enforced through policy instruments at State and institutional levels. Education policy is a discursive instrument which shapes the lived realities of stakeholders within an education ecosystem. Consequently, higher education policy remains key to contemporary human capital formation of which AI has now become an integral consideration. The unpredictable and conspicuous nature of AI presents an added degree of uncertainty with respect to global education practices and how policy might regulate and enhance this. Global South contexts run the risk of undermining its human capital through an unwieldy relationship between AI and its higher education policies.