AI as Digital Capital: New Inequalities in Pakistan’s Higher Education

This blog post, written by Saqib Raza argues that we cannot only analyse the promise of AI as a technology that provides access and improves efficiencies, but also through a sociological analysis of the inequalities it may perpetuate.

Generative AI is often presented as a tool of opportunity. The assumption is simple, once students have access to AI, they will learn better, write better, and compete more fairly. But in Pakistan, this assumption needs to be questioned carefully. AI does not enter a socially equal system. It enters a higher education landscape already divided by class, language, school quality, family background, region, gender, and material resources. In such a setting, AI is not only a technology. It is becoming a new form of digital capital.

Digital capital is not just a device, an internet connection, or an AI account. It is the capacity to turn digital tools into recognized educational advantage. It includes exposure, confidence, language ability, guidance, critical judgment, and the skill to use technology without losing ownership of one’s thinking. Ragnedda (2018) makes this point clearly that digital resources become valuable when they interact with economic, cultural, social, and personal resources. In Bourdieu’s (1986) terms, capital matters because it can be converted into advantage. AI may now become part of that conversion process.

This is why the debate cannot stop at access. Warschauer (2003) argued long before the arrival of generative AI that digital inclusion is not achieved by providing equipment alone; it also depends on literacy, institutions, social support, and meaningful use. The same logic applies today. A student may have access to a free AI tool and still lack the academic confidence, English proficiency, mentoring, and evaluative judgment needed to use it well.

Family background matters deeply here. Education in Pakistan is often more than an individual project. A degree can carry the hopes of a household and may be seen as a route to mobility, security, and respectability. Yet families cannot invest in education equally. Pakistan’s wider human-capital crisis and persistent inequalities in education make these differences especially consequential (World Bank, 2023). Rahman’s (2004) work on education, inequality, and polarization in Pakistan also shows how different school systems are closely tied to socioeconomic stratification.

Elite and educated families often prepare children for academic life long before university begins. Their children may learn how to speak confidently, ask questions, use technology, search for information, and present themselves well. They are guided, corrected, and exposed to academic and digital cultures from an early age. They often have laptops, stable internet, paid subscriptions, English support, and family members who understand the hidden rules of educational success.

Many ordinary students reach university without this support. They may have ambition, but not mentorship; talent, but not exposure; effort, but not the social grooming that helps privileged students move through academic life with ease. The issue is not intelligence. It is preparation. When AI becomes part of academic work, those who already know how the system works are better positioned to turn it into advantage.

English is one of the clearest examples. In Pakistan, English is not only a language. It is a gatekeeping resource tied to prestige, employment, institutional status, and academic confidence. Coleman’s (2010) British Council report documents the central place of English in Pakistan’s education system and the challenges surrounding English-language teaching and learning. Rahman (2004) shows how English-medium, Urdu-medium, and madrassa schooling are connected to social class and educational inequality. This matters for AI because many advanced AI tools work most effectively through English-dominant academic language. Students from elite English-medium schools can often write stronger prompts, judge responses more critically, and polish academic work more easily. Students from Urdu-medium, public-sector, rural, or under-resourced schooling often face a double burden: they struggle with academic English and are then expected to use digital tools shaped by English-dominant knowledge systems.

The rural-urban divide sharpens the problem. Urban elite students are more likely to have better schools, private tutoring, reliable internet, personal laptops, and early digital exposure. Rural and low-income students may face weak schooling, shared phones, poor connectivity, limited study space, and the continuing cost of internet packages. Pakistan’s school pipeline already reflects deep inequality: official education statistics report 26.21 million out-of-school children in 2021-22, and note that children from the poorest households are the most disadvantaged (Pakistan Institute of Education, 2023). ASER Pakistan (2024) also documents learning outcomes and school-level conditions across rural and urban contexts. These inequalities do not disappear at the university gate. They travel with students.

Digital access remains uneven as well. DataReportal (2025) estimated that Pakistan had 117 million internet users at the end of 2025, with internet penetration at 45.6 percent; this also meant that 54.4 percent of the population remained offline. For many students, AI is therefore not simply a click away. Meaningful use requires time, connectivity, privacy, confidence, and a suitable device. Subscription-based AI services may widen this gap further because stronger features become easier to access for those who can pay.

This is why access to AI is not the same as benefit from AI. Two students may use the same tool, but not from the same social position. One may use AI to refine arguments, test ideas, improve structure, and strengthen academic voice. Another may use it mainly to survive deadlines, translate basic text, or cover gaps left by weak schooling. The first student converts AI into academic distinction. The second may become more dependent without gaining deeper learning. The difference is not ability; it is social preparation.

The policy question, then, is not simply whether students should be allowed to use AI. The deeper question is who has the resources to use AI well. Who has the English skills, digital confidence, family guidance, and academic mentoring to benefit from it? Who can afford paid tools? Who is excluded by weak schooling, poor internet, lack of devices, and limited support? UNESCO’s guidance on generative AI in education stresses regulation, human capacity, inclusion, and a learner-centred approach rather than treating AI as a purely technical solution (UNESCO, 2023). If these questions are ignored, AI may become another privilege disguised as innovation.

For Pakistan, the challenge is not simply to introduce AI into higher education. The challenge is to make sure AI does not become a new class advantage. Universities need to support students who lack digital exposure, English confidence, and academic mentoring. AI literacy should therefore include critical judgment, academic writing support, ethical use, language support, and guidance on how to use AI without losing ownership of one’s thinking.

The real issue is not whether Pakistani students use AI. Many already do, and its use will continue to expand. The issue is whether AI will reduce educational inequality or quietly deepen it. If only privileged students can convert AI into academic success, then AI will not democratize higher education. It will create a new digital hierarchy inside an already unequal system.

In Pakistan, AI is becoming digital capital. And like every form of capital, it is not distributed equally. The question is whether universities and policymakers will recognize this early enough, or whether they will allow old inequalities to continue through a new digital form.

The message also goes beyond Pakistan. Many education systems in the Global South face similar conditions; unequal school quality, dominant or colonial language hierarchies, weak digital infrastructure, rural-urban divides, and strong family influence over educational futures. Pakistan is not an isolated case. It shows a wider global problem that AI does not create equal opportunity simply because it is available. Its benefits depend on the social, cultural, linguistic, and economic resources students already possess. For global readers, the Pakistani case is a reminder that digital innovation must be studied through inequality, not only through access, efficiency, or technological promise.

Author bio:

Saqib Raza is an MSc candidate in Educational Leadership in Higher Education at Nazarbayev University, Kazakhstan. His work focuses on generative AI, educational inequality, student autonomy, and higher education governance, especially in non-Western contexts.

References

ASER Pakistan. (2024). Annual Status of Education Report: ASER Pakistan 2023 national report. https://aserpakistan.org/document/2024/aser_national_2023.pdf

Bourdieu, P. (1986). The forms of capital. In J. G. Richardson (Ed.), Handbook of theory and research for the sociology of education (pp. 241-258). Greenwood.

Coleman, H. (2010). Teaching and learning in Pakistan: The role of language in education. British Council. https://apnaorg.com/articles/bc-report/pakistan-language-report.pdf

DataReportal. (2025, November 8). Digital 2026: Pakistan. https://datareportal.com/reports/digital-2026-pakistan

Pakistan Institute of Education. (2024). Pakistan education statistics 2021–22: Highlights report. Ministry of Federal Education and Professional Training, Government of Pakistan. https://pie.gov.pk/SiteImage/Downloads/PES%20Highlights%202021-22%20New.pdf

Ragnedda, M. (2018). Conceptualizing digital capital. Telematics and Informatics, 35(8), 2366-2375. https://doi.org/10.1016/j.tele.2018.10.006

Rahman, T. (2004). Denizens of alien worlds: A study of education, inequality and polarization in Pakistan. Oxford University Press.

UNESCO. (2023). Guidance for generative AI in education and research. https://unesdoc.unesco.org/ark:/48223/pf0000386693

Warschauer, M. (2003). Technology and social inclusion: Rethinking the digital divide. MIT Press.

World Bank. (2023). Pakistan human capital review: Building capabilities throughout life. https://openknowledge.worldbank.org/entities/publication/8748b7a7-7345-4298-9631-3f5f146c7007