EC[ON]OMY

AI’s impact on Kazakhstan’s education system

Artificial intelligence has already entered Kazakhstan’s schools. Not through a carefully approved reform. Not through official standards or detailed regulations. It came through daily practice. Students use generative tools to do homework. Teachers use them to prepare lessons and handle paperwork. Parents use them to check and support learning.

The problem is speed. All this is happening faster than the state can set clear rules. This gap between reality and governance became the key finding of a large study conducted in 2025-2026 by the TALAP Center for Applied Research. More than 300 representatives of the education system took part.

The main message of the study is simple and uncomfortable. The question is no longer whether AI should be introduced into schools. It is already there. The real question is different – who controls this process, and how. Kazakhstan has reached a turning point. The cost of mistakes here is not a failed IT project or wasted money. It is the long-term quality of human capital.

From the start, the study sets a clear frame. AI in schools is not a technology project. It is a social and institutional shift. Mistakes show their impact years later. Global experience confirms this. No country has yet built a stable, finished model of AI integration in school education. Everyone is moving through pilots, pauses, revisions, and sometimes rollbacks. Kazakhstan is in the same zone of uncertainty, but with its own deep structural problems in education that AI does not fix. In many cases, it makes them worse.

The diagnosis is direct. AI is widely used, but in an unstructured way. It enters a system with weak functional literacy, formalized assessment, and declining teacher authority. In this setting, AI does not raise quality. It hides weaknesses.

A student may not think independently, but if they can write a good prompt, AI produces a polished answer. If assessment checks form instead of reasoning, the problem stays invisible. On the surface, progress looks real. Inside, the system slowly degrades.

The study highlights another key issue. AI entered schools before education goals were updated. The classic chain school – university – profession is breaking down. Automation now affects not only routine tasks but complex intellectual work. Career paths are no longer linear or stable. They are fragmented and project-based. In this reality, the old promise of school no longer works. Passing on a fixed set of knowledge no longer guarantees a predictable future.

Uncertainty is no longer temporary. It has become the new normal. AI development follows several competing paths. One leads to mass automation. Another leads to new industries and new forms of work. No single future can be chosen. Education policy must work across multiple scenarios at once.

This changes the role of school itself. With unlimited access to information, simple knowledge reproduction loses value.Schools face a real choice. One path leads to discipline and control. The other turns school into a space for thinking, personal growth, and cooperation. Participants in the study clearly see the second path as socially necessary. But it requires a different governance model and a different role for teachers.

The teacher’s role does not shrink. It becomes more complex. Authority based on knowing more than students fades. What matters instead is guiding learning paths, developing thinking, supporting motivation, and encouraging reflection. These tasks cannot be automated. That is why the study stresses that without serious investment in teachers, any AI initiative will turn into imitation rather than real change.

The study also outlines the risks of uncontrolled transformation. These include weaker thinking skills, lower independence, and a shrinking range of learning strategies. The biggest concern is early grades. Even second and third graders already delegate basic thinking tasks to AI. This strengthens the case for protecting core skills. Handwriting, mental math, memory, fine motor skills – this is not nostalgia. It is about protecting the cognitive foundation.

Kazakhstan’s school reality is described without illusions. Low functional literacy is not a problem of individual schools. It is systemic. Official documents talk about skills and competencies, but real outcomes remain weak. Assessment often serves reporting needs rather than feedback. In this environment, AI does not solve problems. It locks them in.

At the same time, alternative education markets are expanding. Tutors, exam prep centers, online platforms are filling the gap. This reflects a crisis of trust. Schools are no longer seen as places that guarantee real learning results.

The study analyzes the positions of all key stakeholders. Students, parents, teachers, school leaders, methodologists, government officials, and EdTech companies agree on one thing. AI is unavoidable. But they disagree on boundaries, speed, and purpose. These conflicts cannot be solved by orders. They must be managed.

There is also a clear area of consensus. It is not about technology itself, but about conditions. Updating curriculum content. Moving from formal testing to project-based and portfolio assessment. Prioritizing psychological safety. Retraining teachers. Building data and infrastructure capacity. This is the foundation for gradual, realistic policy.

Global experience supports caution. After 2023, many countries shifted focus from AI benefits to managing consequences. OECD and UNESCO describe school AI as a field of ongoing experiments, not ready-made solutions. South Korea slowed down after public backlash. China tightened regulation. Singapore focused on responsible use. The EU set legal limits. One conclusion stands out. Leadership today is not about speed. It is about balance.

For Kazakhstan, this means there is no ready template to copy. The study states this directly. Imported models do not work. The country must build its own governance logic. Two paths are possible. One allows market forces and user practices to drive change. The other builds a phased, controlled, and legitimate transformation. The document clearly supports the second path.

Project-based approaches are proposed as a practical tool. They turn chaotic use into managed pilots, reveal risks, and generate data. More than 50 AI-related school initiatives were identified during the study. They are not final answers. They are material for policy design.

At the institutional level, the study proposes professional associations, structured parliamentary discussion, and systems to monitor risks and effects. The goal is to close the gap between the state, the market, and schools.

The document avoids both techno-optimism and fear. This is an adult conversation about limits, risks, and responsibility. Artificial intelligence in Kazakhstan’s schools is framed not as a quick reform, but as a long process. The key asset remains the human being. The key risk is losing control of the system.

Lina Yegil kizi, expert of the EconomyKZ.org portal

Scroll to Top