IMPACT OF AI IN EDUCATION AND DIGITAL INEQUALITY AMONG SOCIAL GROUPS IN KAZAKHSTAN

Abstract. Artificial intelligence (AI) is changing education systems around the world with its rapid development. AI-powered tools and platforms are opening up new opportunities for personalized learning and expanded access to learning resources. However, new innovations are also introducing deep issues of equity and access: most notably in countries like Kazakhstan, where digital infrastructure and technology adoption are uneven across urban and rural (village) environments.

The aim of this study is to examine the gap between the adoption of AI in education in urban and rural areas of Kazakhstan. While students in urban areas may enjoy more modern devices, faster internet, and their teachers are already integrating AI into their learning, rural students are faced with poor technology infrastructure, such as slow internet, and their teaching methods remain outdated. By comparing teachers' and students' experience with using AI in both environments, the research will investigate how AI reduces or expands gaps in education.

Background.

Despite Kazakhstan’s remarkable achievements within last years in technology development, digital inequality continues to limit inclusive access to technological innovation in key areas such as education. While the government’s Digital Kazakhstan program has increased internet access to 92%, this number does not tell the full story in a country where many rural areas have poor or no internet connection. Rural regions, which account for nearly 40% of residents, will probably experience patchy or slow connections despite proclaimed broadband reach. For instance, in most parts of the country, internet speeds had been stuck at 20 Mbps until 2021, and fiber-optic networks still needed to extend to distant villages. Kazakhstan's immense geographical area, over 2.7 million square kilometers, makes it an enormously expensive undertaking to lay telecommunications infrastructure in inaccessible regions, resulting in a largely uneven territorial divide in access to digital services.

In addition to infrastructure, digital literacy is also a major obstacle. In 2021, the lowest regional rate of digital literacy among the population aged more than six was registered in the Akmola region at a mere 68.9%, against the country-wide average of 79.6% (Kurmanov et al., 2022). While digital literacy is expanding throughout the country, the uneven nature of development still constrains equal access to AI-facilitated learning, especially in less developed regions. Socioeconomic limitations and lower levels of digital literacy, especially in rural communities, further exclude those latter groups from taking advantage of AI-facilitated learning technologies. High-speed internet access remains largely centralized in cities such as Almaty and Astana, while rising mobile tariffs and inflation-driven cost pressures disproportionately harm low-income households (Kurmanov et al., 2022). This vision is one that implies education AI will further exacerbate inequality. It will deliver quality learning to those already possessing good technology, internet, and device access, but further exclude those already disadvantaged based on where they reside, how much income they earn, or simply because they do not possess the required technology and internet. Any model of incentive-driven AI application in education in this vision will thus have to meet such exclusionary disparities existing in Kazakhstan's digital space with stringent rigor

Discuss how the application of artificial intelligence in education has impacted social equity in Kazakhstan, and whether it can be attributed to prevailing societal issues such as digital inequality, disparities in access to technology across various regions, and marginalization of rural or underprivileged student groups.

Method Research:

The research approach employed in this study is a mixed-method one involving qualitative and quantitative approaches. Data used in this study are qualitative data such as answers to interviews and text analysis, and quantitative data gathered through the use of questionnaires. Data sources are collected from library work, using online sources such that scientific articles and reputable news sources are utilized. The data sources are categorized in themes and synthesized using a comparative analysis method. Literature review, interviewing of major stakeholders (e.g., teachers and students), and a quantitative questionnaire administered to the wider population are methods of data collection used in this study. All data gathered are logically examined to determine patterns and observations, against which the foundation upon which to draw inferences regarding fairness as it pertains to artificial intelligence is developed.

Literature review:

In Digital Divide of Resource-Based (Oil and Gas) and Service-Dominated Regions, a paper in the Journal of Open Innovation (2022), Kurmanov et al. explore how Kazakhstan's regional economic orientation differences influence digital infrastructure and access. Research finds that those regions dependent on extractive industries lag behind those service-dependent ones in terms of digital progress, leading to an unequal availability of high-speed internet, which affects access to digital services like education. They highlight that mobile broadband speeds in 2021 were around 20 Mbps in the majority of rural communities, and fiber-optic rollouts are low, especially in remote settlements. This is a geographical disparity that is an infrastructural challenge for AI-based learning tools, which rely heavily on stationary, high-bandwidth internet connections.

Affirming this view, the 2024 USAID Digital Ecosystem Country Assessment (DECA) concludes that while Kazakhstan has high national levels of internet penetration, rural communities and disadvantaged groups suffer disproportionately. The assessment notes that the lack of strategic digital literacy planning coupled with linguistic and socio-economic disparities results in systemic digital education ecosystems exclusion—especially by rural youth, ethnic minorities, and women.

The UNDP (2023) has highlighted persistent gendered and geographical access barriers to digital technology where rural women in particular have gaps in social norms and digital competency to contain their participation in technology-enabled learning spaces.

Despite the rising global literature on the education role of AI, local implementation in Kazakhstan has not been given sufficient attention. Earlier work (e.g., 2010–2020 meta-analysis of AI and learning) positions AI tools as three generations: rule-based systems, statistical learning, and brain-inspired neural models. But these studies generally presume stable digital access and disregard the ways in which uneven infrastructural realities shape user experience and results. Not much has been spent on understanding how these AI tools operate in the context of the digital divide.

While the existing literature makes positive contributions to understanding the digital divide and the development of AI in education, it does not offer a detailed examination of how these dynamics play out in fast-growing economies like Kazakhstan. Some key questions remain:

1. How does unequal access to AI-based educational technologies affect student learning outcomes in urban and rural areas of Kazakhstan?

2. How do teachers and students in Kazakhstan perceive the benefits and challenges of integrating AI tools into the classroom, and how do these perceptions vary across regions?

 3. To what extent does the growing use of AI in education in Kazakhstan reinforce or reduce existing digital inequalities among students?

This study addresses these lacunae by comparing urban and rural learning spaces in Kazakhstan with an aim of creating a model of digital equity that critically examines who benefits from AI in education and who is systematically excluded.

Results and Discussion

Survey data were collected from 120 students (60 urban and 60 rural). The findings reveal clear disparities:

• Access to Devices: 82% of urban students reported having personal laptops or tablets, compared to only 47% of rural students. For rural respondents, smartphones served as the primary device for 76%, which often limits the functionality needed for advanced learning platforms.
• Internet Quality: 74% of urban schools reported “stable and fast” internet access, while only 28% of rural schools indicated the same. Furthermore, 42% of rural students reported frequent internet disruptions during class, making it difficult to use AI tools consistently.
• Use of AI Tools: 65% of urban students said they used AI-based tools (such as ChatGPT, or Grammarly AI) at least once per month, whereas only 23% of rural students reported the same.
• Perceptions of AI: 85% of urban students agreed that “AI makes learning easier,” compared to only 43% of rural students.

Qualitative interviews with two teachers (one urban, one rural) of the same subject add depth to these findings:

• The urban teacher reported successfully integrating AI into lesson planning and assessment. They described AI tools as effective for generating quizzes, providing instant feedback, and supporting differentiated instruction for students at varying levels.
• The rural teacher, by contrast, highlighted major obstacles: insufficient training in AI tools, weak infrastructure, limited access to reliable devices, and in some cases resistance from the school administration to adopt “new” technologies.

Analysis:

These results indicate that the adoption of AI in education is highly dependent on infrastructure and teacher readiness. Urban environments provide both the technical foundation (fast internet, devices) and cultural acceptance for AI integration, whereas rural environments continue to lag behind. Importantly, the gap is not only technological but also perceptual: rural students and teachers tend to view AI with skepticism, largely because their experience of it is interrupted by technical limitations. This creates a cycle where lack of exposure reinforces negative perceptions, further discouraging adoption.

Overall, the findings suggest that AI has dual potential: it can enhance equity by providing personalized learning opportunities, but without equal access, it risks amplifying existing inequalities between rural and urban education systems.

Conclusion. This study demonstrates that digital inequality in Kazakhstan is a decisive factor shaping how AI is integrated into education. The survey data revealed significant disparities: urban students benefit from better device ownership (82% vs 47%), more reliable internet (74% vs 28%), and higher usage rates of AI tools (65% vs 23%). At the same time, qualitative interviews revealed contrasting teaching realities: one teacher successfully integrates AI into pedagogy, while the other struggles with infrastructural and institutional barriers.

Key insights include:

1. Infrastructure as the foundation: Without stable internet and access to modern devices, rural schools cannot meaningfully adopt AI.
2. Teacher training and support: Even where infrastructure exists, lack of professional development prevents rural teachers from using AI effectively.
3. Perceptions and trust in AI: Rural communities often view AI with skepticism due to limited exposure and technical failures, which hinders acceptance.
4. Policy implications: Government investment must go beyond connectivity statistics (e.g., 92% national internet access) and target quality of service in underserved regions.
5. Equity-driven design: AI in education should be localized for Kazakh contexts, including language support and offline capabilities, to serve students in low-connectivity areas.

By addressing these systemic issues, Kazakhstan can transform AI in education from a driver of inequality into a tool for digital fairness. The long-term vision should be an education system where every student—regardless of geography—benefits equally from technological innovation.

Appendix A: Survey Questions and Results

Appendix B: Teacher Interview Transcripts

Appendix A: Survey Results

The following chart compares urban and rural students' responses on device access, internet stability, AI usage, and perception of AI in education.

Figure 1. Results

Appendix B: Teacher Interview Results

The chart below summarizes the responses from two teachers (urban vs rural) regarding their experiences with AI in lesson planning, assessment, perceived benefits, and infrastructural challenges.

Figure 2. Results