AI in Central Asia: How to Avoid Stepping on Others' Pitfalls and Build Your Own Technological World
In the first article of our series – "AI in Central Asia: So, Who Is This Multifaceted Artificial Intelligence?" – we examined how perceptions of the technology have transformed, from kitchen-table myths and fears to real automation tools in the economy. Now it's time to dive into practical applications and see how these solutions are taking root in our region.
The five Central Asian states have entered the era of artificial intelligence with varying levels of digital maturity. This initial capital has not only predetermined their current successes but also shaped specific challenges that some of their regional neighbors have already encountered, while others have yet to.
I am confident that understanding these differences and the emerging conditions is important both theoretically (to identify hidden patterns and vulnerabilities) and practically, so that, drawing on the experience of our regional colleagues, we can quickly avoid the mistakes they have already made. In other words, avoiding the same mistakes as those who have already made.
Different Starting Positions
By the time smart algorithms began to take off, Kazakhstan had already cemented its status as a technological leader in the region. Its main advantage was the solid infrastructure foundation of the "Digital Kazakhstan" state program and a developed e-government ecosystem, which allowed it to collect and organize colossal amounts of data.
A special role was played by the Astana Hub international technology park, opened in 2018 on the infrastructure of the previous EXPO, as well as a strong school of young engineers, a stable telecommunications infrastructure, and the availability of high-speed internet.
Uzbekistan experienced the wave of artificial intelligence (a true "ninth wave," to use Aivazovsky's example) in the midst of large-scale economic and technological modernization. The foundation for this breakthrough was a radical reform of the IT sector, the opening of the innovative IT Park in Tashkent, and a strong demographic dynamic with a clear predominance of young people.
But the fundamental educational initiative, in our view, was the "One Million Programmers" project. This ambitious undertaking not only proved successful but also exceeded its initial goals. The total number of registered users has exceeded 2.5 million (!), of whom more than 1.1 million successfully completed the training and received certificates.
Kyrgyzstan chose the path of a flexible, adapted digital environment, the drawback of which was inherent in its decentralization. However, a strength was the implementation of the X-Road e-government data exchange architecture through the national Tunduk platform.
However, from the very beginning, the country lagged behind regional innovators in terms of financial resources and the stability of its energy infrastructure, which is extremely critical for the smooth operation of computing systems and data centers.
Tajikistan began in more challenging conditions: low internet penetration and the dominance of the agricultural sector forced their own adjustments. Most likely, this is why the country's leadership relied on the concept of leapfrogging, recognizing that without modern algorithms, the country risked falling into the trap of technological backwardness.
Turkmenistan developed under a centralized governance model. As part of the State Program for the Development of the Digital Economy, the country carried out a comprehensive modernization of telecommunications, digitalization of the public administration system, and the implementation of IT education programs for government officials and ordinary citizens.
With the support of the Islamic Development Bank, projects totaling $273 million were implemented to expand and modernize telecom operator networks, achieving 100% high-speed internet coverage in populated areas.
A striking example of the practical integration of technologies is the Arkadag smart city, the first "smart city" in the region. All its residential, administrative, and industrial facilities are equipped with smart home systems, and artificial intelligence-based solutions are integrated into the control and security systems.
Furthermore, in preparation for the new technological era, Turkmenistan has updated its regulatory framework for personal data protection, strengthened cooperation with international partners on cybersecurity, and launched educational programs on digital hygiene for all age groups.
The Current Status of Technologies – From Declarations to Institutions
In recent years, the countries of Central Asia have undergone a qualitative transition – from declarations and general statements of intent, the region has moved toward the creation of specialized agencies and the adoption of a specific regulatory framework.
For example, Kazakhstan has experienced an ecosystem breakthrough in public administration – the country has successfully established a Ministry of Artificial Intelligence and Digital Development, and a specialized law has been adopted. At the infrastructure level, in addition to the Almaty computing cluster, a new powerful national supercomputer focused on complex neural network computations is being deployed. A National AI Platform has been created.
One of the most important technological achievements, in our view, was the AlemLLM language model in Kazakh, which is open to business. Therefore, it is not surprising that, according to international estimates, Kazakhstan is firmly among the top thirty global leaders in per capita use of generative neural networks.
Uzbekistan is also demonstrating a flexible, pragmatic approach. The country is effectively implementing the "Strategy for the Development of Artificial Intelligence Technologies until 2030." Legislation has enshrined the legal definition of AI and explicitly prohibited automated decisions that infringe on human rights.
This is a crucial precedent for the entire region, clearly establishing that individuals have the legal right not to be subject to decisions made by algorithms without their personal participation. Otherwise, this could lead to undesirable legal or financial consequences.
In practice, more than twenty major projects have been implemented in Uzbekistan, including the MyID and FaceID biometric systems in the banking sector, palm-based payments in the Tashkent metro, the LexAI legal platform, and smart assistants in the Tax Committee, among others.
Kyrgyzstan has chosen a path of relaxed regulation and support for the IT community. Instead of rigid legislative frameworks, the emphasis is on self-regulatory organizations and ethical codes to give businesses the freedom to experiment. At the same time, the country is developing a network of modular data centers.
Internationally, Bishkek has launched an initiative to create a Regional Artificial Intelligence Hub – a major technological platform for ensuring digital sovereignty, integrating AI into government processes, and creating a unified, independent network ecosystem within the Eurasian Economic Union.
Tajikistan has achieved significant geopolitical positioning – the United Nations General Assembly adopted a Resolution on Artificial Intelligence, initiated by Dushanbe. Furthermore, a Regional AI Center has been established in the capital under the auspices of the UN, aiming to coordinate programs in Central Asia. In Tajikistan's financial sector, smart algorithms are already used to evaluate approximately a quarter of all loans issued.
Turkmenistan continues to systematically develop the Arkadag smart city, where an intelligent transportation system is currently being actively implemented and utility networks for the city's second development phase are being automated.
Targeted work is also underway to establish an institutional framework for the development of AI. The adoption of significant documents—the National Concept for the Development of Artificial Intelligence and a detailed Implementation Plan—is on the agenda.
Undoubtedly, these steps by Turkmenistan, aimed at establishing a systemic approach to technology implementation, will allow for the synchronization of efforts to develop artificial intelligence among ministries, agencies, and academic and university circles in the near and medium term.
Where AI Is Used and How the Industry Is Supported
The main areas of practical application of artificial intelligence in Kazakhstan are concentrated in e-government, fintech, education, and the oil and gas sector. The government supports startups by providing them with the computing power of the national supercomputer and organizing large-scale courses for hundreds of thousands of students.
In Uzbekistan, priority areas are financial technology, tax administration, transport biometrics, and public services. Support measures include targeted grants and a network of specialized research laboratories.
Kyrgyzstan is focusing on IT outsourcing of AI services, digital government services, and the localization of language models. The government is stimulating this process by building data centers and providing graphics processing power to small businesses.
In Tajikistan, priority is given to bank scoring (credit assessment), resource management, and agriculture. Development is supported by international grants and the work of the specialized AI Council.
Turkmenistan is currently in the process of industrial analysis and planning. The "Concept for the Development of the Digital Economy for 2026–2028" has been adopted, officially acknowledging artificial intelligence as a driver of economic growth. At the departmental level, a systematic study is underway to integrate smart algorithms into critical sectors such as the oil and gas sector, geology, water resources management, logistics, and others.
Understandably, special emphasis is placed on ensuring cybersecurity (Turkmenistan has established a specialized State Service for this purpose), optimizing energy grids using neural network solutions, and consistently automating public administration through the expansion and modernization of the national data transmission infrastructure.
Prospects and Main Directions
It can be assumed that several vectors of artificial intelligence development will dominate in Central Asia in the coming years.
The first direction is the struggle for linguistic and cultural sovereignty. AI specialists in Central Asian countries (and elsewhere) have come to a clear understanding that global English-language models are incapable of fully reflecting the local context, history, and specific features of national languages. Therefore, the creation of their own sovereign language models will become a priority for many states, including Turkmenistan.
The second vector is overcoming the shortage of computing power and the energy needed to operate it efficiently. The main constraint on the development of artificial intelligence in the region will not be a lack of ideas or algorithms, but rather a growing shortage of specialized equipment, electricity, and sufficient water to cool large data centers.
The third mainstream is institutional integration. In the future, countries in the region will strive to coordinate their efforts, while international initiatives will require adherence to uniform standards for cross-border data exchange.
If we rank countries by their level of readiness for the upcoming steps, Kazakhstan has a fully-fledged triad: a relevant law, a supercomputer, and a national language model.
Uzbekistan, it's safe to say, is also at a breakthrough level. The country is demonstrating a relatively high rate of practical implementation of AI in the real economy and public administration.
Kyrgyzstan is adaptive – it has a strong IT community, but budget constraints are a limiting factor.
Tajikistan has established itself at the initial, progressive stage, where it demonstrates significant international positioning despite a modest technical base. However, progress is evident. For example, at the Central Asia – Republic of Korea Think Tank Forum held in June of this year, an initiative to establish a Tajik-Korean Digital Competency Center was announced.
In turn, Turkmenistan, among its priorities for the upcoming UN General Assembly session, proposes developing, under the auspices of the UN, an International Code for the Peaceful Use of AI in International Relations and Diplomacy, as well as holding an Expert Dialogue in the Turkmen capital, with the agenda focusing on digital trust and the responsible use of artificial intelligence.
Turkmenistan is relying on a gradual, systemic approach to building its own trajectory, first establishing the foundation and rules, followed by scaling and integration. This tactically balanced approach is entirely justified.
Similar principles of consistent implementation can be found not only in our region but also in other countries where technological development initially proceeded through local solutions and strict security frameworks.
In China, for example, the development of AI infrastructure relied for a long time on its own cloud and computing platforms, with limited access to foreign services. Over time, this approach spurred the creation of strong domestic developments and standards, which later became globally competitive.
In a number of government projects, India prioritized the creation of a trusted environment over the rapid implementation of newfangled models—gradual data verification, ensuring algorithm transparency, and integrating AI into government services without losing control over quality and security.
In the United Arab Emirates, the AI development strategy was built through targeted pilot projects and national roadmaps, where every step was pre-agreed with legal norms and ethical frameworks.
In all these cases, a focus on manageability and a gradual approach helped avoid systemic risks and build a sustainable AI ecosystem.
Preliminary Summary
The AI landscape in Central Asia is currently highly dynamic. However, the region as a whole is still playing catch-up with global technology leaders, whose experience requires deep analysis.
At the same time, Central Asian states must maintain a reasonable balance when interacting with foreign vendor developers to avoid the risk of critical technological dependence ("technological colonialism") and preserve their digital sovereignty.
The main systemic risk for the region remains market fragmentation. If Central Asian countries fail to develop common standards, mechanisms for mutual data recognition, and the integration of compatible digital platforms over time, their domestic markets will remain too localized to develop full technological autonomy.
Of course, the range of challenges and risks associated with the implementation of artificial intelligence technologies is quite broad – and a detailed analysis of key security threats, legal conflicts, and ethical aspects will be the subject of a separate article.
(To be continued)
Bekdurdy AMANSARYEV,
Center for Strategic Studies, Institute of International Relations, Ministry of Foreign Affairs of Turkmenistan
On this topic:
"AI in Central Asia: So Who Is This Multifaceted Artificial Intelligence?"
