It is the Asian "IT tiger". It has already overtaken Poland in an important ranking
- Artificial intelligence is the next stage of technological progress and more and more countries around the world are declaring their interest in having their own models.
- Kazakhstan ranked 34th out of 67 countries included in the study in digital competitiveness. Poland came in 39th.
- AI (Artificial Intelligence) is not only an opportunity to accelerate economic development, but also a number of challenges for the labor market, information security and society, and counteracting them will require coordinated actions in the legal, scientific, business and social spheres.
Kazakhstan has been systematically implementing the process of digitalization and development of IT infrastructure and the IT sector for several years. Thanks to this, it has become a regional leader in this area and a model for other republics. The high dynamics of the republic's ICT development was undoubtedly contributed to by the Digital Kazakhstan program , announced in 2017 by President Nursultan Nazarbayev, which assumed increased spending on the digitalization of the economy, the development of the IT sector, infrastructure and public services available online.
In 2024, the digital competitiveness index ( IMD World Digital Competitiveness Index ) of the Swiss Institute for Development Management (IMD), which determines the state's ability to use digital technologies for economic growth, amounted to 66.43 points out of a possible 100 for Kazakhstan . This gave the republic 34th place out of 67 countries included in the ranking. Poland was ranked 39th out of 63 points. In turn, in the UN ranking in the area of digitalization of public administration services, Kazakhstan took 1st place in the Commonwealth of Independent States and 24th in the world.
Astana's achievements in computerization are confirmed by the connection of over 1,200 rural areas to the ICT network and the provision of access to fast connections for almost four thousand state institutions. Currently, over 93 percent of public services are available to citizens online, including 86 percent adapted to mobile devices. Kazakhstan also has the most developed e-commerce network in Central Asia, as well as B2B transaction platforms. According to data from the National Bank of Kazakhstan, the number of cashless transactions, including those carried out via online banking and mobile applications, is systematically growing. In March 2025, they accounted for 79.8 percent of all cashless transactions, and their share in terms of value was 89.3 percent.
In 2024, Kazakhstan took another step towards IT development and strengthening its position as a regional leader in this sector, with the announcement by the Ministry of Digital Development of the strategic AI development plan for 2024-2029, which aims to create a platform for the development of artificial intelligence and stimulating innovation.
In January 2024, the International Monetary Fund (IMF) published the report “Gen-AI: Artificial Intelligence and the Future of Work.” Kazakhstan has the highest AI Preparedness Index (AIPI) among the Central Asian republics (0.55) , significantly exceeding the average for the region (0.4) and approaching the EU average (0.66). The republic achieved this high index thanks to its AI development plan, the creation of the Alem.AI international startup center, and the Tomorrow School AI training program and TUMO Education, a network of digital centers developed mainly in Armenia by the Armenian diaspora, which recently opened in Kazakhstan.
President Kassym-Jomart Tokayev also announced that he would conduct legislative work on AI in cooperation with foreign experts. The AI Development Council is to serve this purpose. The process of implementing AI-based tools is already advanced in the judicial system and is used to analyze cases and predict verdicts, which is to help reduce the number of miscarriages of justice. Also in the health service, since 2022, the PneumoNet pilot program has been operating, which allows for the detection of 17 most common lung pathogens, including pneumonia, tuberculosis and cancer.
AI development requires expensive infrastructure, including supercomputers. In March 2025, President Tokayev met with Thomas Pramotedham, head of Presight AI from the United Arab Emirates (UAE), which develops artificial intelligence. The topic of the talks was the creation of a supercomputer cluster .
Geopolitics of new technologies creates challengesThis essential stage of work on AI has now become a serious challenge for Kazakhstan, not because of costs or technological problems, but because of geopolitical changes. Kazakhstan's computer systems have so far been based on Russian systems , but Russia does not have its own AI processors, and in the face of sanctions imposed by the West, it has very limited access to innovative technologies. The turn towards the UAE company also means reaching out to the Americans, because the supplier of processors for the construction of the supercomputer is NVIDIA, which controls 80 percent of the market for these components.
And here a problem arose, because Washington imposed export restrictions due to the risk of Moscow taking over the technology. The construction of the supercomputer, which was supposed to be put into operation by the end of last year, has been stalled for an unknown amount of time. This may prompt Astana to seek help from China, because Huawei has a suitable Ascend 910C processor , although not as efficient as NVIDIA's product.
How to create your own language model (LLM)?The second challenge for Kazakhstan, apart from infrastructure, is the large language model (LLM). The most popular models are ChatGPT, Lama and Rok, which process and enable understanding of the delivered content and generate human speech in many languages, but do not include Kazakh.
In order to create its own national product, Kazakhstan must have its own model. Additionally, global models are characterized by "Western-centricity", i.e. perception of reality through the prism of the Western world's mentality. Therefore, unique cultural or mental aspects of other nations, outside the circle of Western culture, are not taken into account. This problem has already been solved by, among others, the Chinese, who created their own DeepSeek model, or the Taiwanese, who have Taide. However, these are "Sinocentric".
The development of LLM itself is only the beginning of the process, because it is necessary to provide the model with huge amounts of data, i.e., to ensure an effective learning process. And here another problem arises, which is the relatively modest amount of data on the web. According to Common Crawl, which is a source of data for teaching AI, only 0.0298 percent of websites are in Kazakh . Additionally, search engines first choose English-language websites, AI assistants have problems with queries in less popular languages, and automatic translations are often inadequate for the content sought. This problem concerns, among others, the Kazakh language, which is translated via Russian.
However, Kazakhs are close to overcoming this barrier. In December 2024, KazLLM was presented at the Institute of Intelligent Systems and Artificial Intelligence at Nazarbayev University. The second model, Sherkala, based on Llama, was presented in February 2025 at the Mohamed bin Zayed University of Artificial Intelligence (MBZUAI) in Abu Dhabi. Both take into account the specific cultural and mental characteristics of Kazakhs and are aimed at collecting data on the history and culture of the republic.
New opportunities and new threatsThe implementation of the AI platform would likely be an impulse for the dynamics of Kazakhstan's economic growth, would increase the republic's investment attractiveness and its competitiveness in Asia.
On the other hand, each country implementing AI must face the difficult-to-define problem of the impact of this technology on the structure of the labor market and the threat of its use for illegal activities. In the case of developing countries, such as Kazakhstan, the risk of a strong negative impact on the labor market is relatively low, but the threat of digital exclusion among the elderly and poorly educated is high . Therefore, the implementation of AI must be accompanied by a strategy for adapting society to new conditions with the widespread presence of AI in all areas of life.
Another area of challenges is ensuring cybersecurity, especially financial transactions and the flow of sensitive data. Another problem is creating a legal framework for the functioning and use of AI, which, on the one hand, will penalize certain behaviors, and on the other hand will not restrict the development and use of AI. Another task for the government is a broad social education program, the aim of which will be to create awareness of the existence and willingness to use new tools among citizens.
However, for this purpose it is necessary to guarantee the safety and real benefits of using AI tools, and this task requires coordination of the activities of the government, local administration, business and the education system. And this may turn out to be more difficult and complicated than the process of creating an AI platform.
wnp.pl
