Speaking on October 1 at the AI & Digital Bridge forum in the capital city of Astana, Kazakh President Kassym-Jomart Tokayev proposed an AI Challenge focused on solving industry-specific problems. The initiative invites both local and international teams to compete for solutions, with the most effective projects moving into actual production. This proposal highlights the core dilemma facing Kazakhstan’s technology agenda: converting digital capabilities into tangible economic returns.
This objective is critical for a nation striving to diversify its economic drivers beyond natural resources. Technology provides a dual pathway toward this goal. While software companies can build export streams, artificial intelligence can enhance the productivity of existing sectors. For Kazakhstan, applying digital tools to minimize equipment downtime, optimize freight operations, and streamline bureaucracy gives its technology strategy immediate economic relevance.
Kazakhstan possesses a solid starting point. The country ranked 24th out of 193 nations on the 2024 UN E-Government Development Index, securing the tenth spot globally for online service delivery. Transitioning government functions online creates a natural springboard for integrating AI into tools citizens already navigate daily. The upcoming challenge involves making those digital portals more intuitive and tailored to individual requirements.
Evidence of this shift is already emerging. Government data from August showed that the eGov GPT assistant provided access to 50 public services and served upwards of 150,000 residents. Additionally, digitizing high-demand services cut physical visits to public service centers by 30%. These outcomes demonstrate why artificial intelligence adoption benefits from prior groundwork: the necessary administrative structures and user base are already established.
Following President Tokayev’s declaration of 2026 as the Year of Digitalisation and Artificial Intelligence, authorities rolled out a comprehensive policy structure. The Digital Qazaqstan strategy received approval in June and spans through 2029. Its accompanying August execution plan outlines specific metrics linking tech budgets to results, such as halving the processing time for public services and social welfare programs, alongside hooking up a minimum of 70% of energy and fuel infrastructure to digital monitoring networks by 2029.
Although these remain objectives for now, they provide a framework for tracking advancement. The same performance-oriented logic should apply to the planned industrial AI Challenge. A predictive maintenance platform must prove it lowers equipment failures, while a logistics tool needs to demonstrate faster cargo transit. Competitions deliver real economic value only when winning prototypes successfully win over clients and scale past pilot phases.
Connecting infrastructure with local expertise
Hardware infrastructure forms another pillar of the strategy. During the conference, Tokayev’s discussion with NVIDIA Vice President Rev Lebaredian centered on two operational supercomputing clusters running on NVIDIA hardware, alongside plans for Data Center Valley in Ekibastuz, a proposed expansion of regional computing capacity. This distinction is vital: current assets handle immediate workloads, whereas large-scale projects demand long-term capital and steady power supplies.
The true worth of these systems relies on the applications built by domestic researchers and enterprises. KazLLM, created by Nazarbayev University’s Institute of Smart Systems and Artificial Intelligence, demonstrates the significance of home-grown expertise. Built on the Llama 3.1 architecture, the model was engineered to improve processing in the Kazakh language and incorporate regional nuances. Tailoring foundational technology helps resolve specific local demands often overlooked in mainstream global AI research.
Language proficiency carries practical implications. Public tools become far more useful when citizens can interact using their native tongue. Educational software similarly depends on clear, accurate native explanations. A raw language model alone is insufficient; it requires integration with verified data, rigorous testing, and continuous iteration. Kazakhstan’s opportunity lies in cultivating this technical know-how alongside its hardware investments.
The newly launched Qazaq AI Research University introduces a dedicated academic pillar. Its curriculum spans artificial intelligence, machine learning, physical AI, and AI+X, which merges AI studies with other academic fields. Access to national computing clusters can help students transition directly from the classroom to practical research. Meanwhile, the national strategic roadmap targets basic AI literacy for at least 80% of graduates from schools, colleges, and universities by 2029.
Turning capabilities into businesses
Commercial indicators show the tech ecosystem is maturing. The Kazakhstan AI Country Report, compiled by RISE Research in collaboration with stakeholders like Mastercard and Freedom Bank, tracked more than 100 artificial intelligence startups. The study estimates that venture capital funneled into AI initiatives grew from roughly $14 million in 2023 to $73 million in 2025. Many of these firms concentrate on enterprise applications, allowing buyers to evaluate solutions against explicit operational metrics.
A prominent example is Higgsfield. The AI video enterprise maintains its headquarters in San Francisco alongside a major engineering division in Almaty. In August, the firm secured a $400 million funding round at a valuation of $5.4 billion. Its expansion illustrates how developers based in Kazakhstan can contribute to internationally marketed goods, bridging domestic talent with global financial markets and consumer bases.
Cross-border demand is also on the rise. Tokayev informed attendees at Digital Bridge that IT service exports from Kazakhstan surpassed $1 billion in 2025. Export growth grants local tech creators opportunities outside their domestic borders. Firms that refine their offerings against the demands of international clients can subsequently channel that expertise back into domestic initiatives.
Achieving broader scale will require overcoming distinct hurdles. The AI Country Report points out persistent gaps in quality industry datasets and specialized talent, regional disparities in digital infrastructure, and a scarcity of late-stage venture financing. These limitations clarify the heavy focus on educational reform and corporate partnerships. Furthermore, enterprise clients require assurances that incoming technologies will receive ongoing maintenance and technical support post-deployment.
Regulatory frameworks will heavily influence that trust. Although Kazakhstan has passed an AI law and a Digital Code, Tokayev cautioned the audience: “We must not overregulate this sphere to the point where promising teams, technologies and capital begin moving to other jurisdictions.” Striking this balance involves safeguarding user data and establishing operational transparency while giving creators the freedom to test innovative tools. Citizens must also retain avenues to contest critical errors stemming from automated systems.
Kazakhstan’s methodology merits close observation because it ties state-level AI aspirations directly to established public digital platforms and practical economic requirements. Success will ultimately be measured by execution: whether administrative platforms save citizens time and whether businesses secure real productivity gains. Delivering these results would cement a robust domestic technology sector and provide other emerging markets with a blueprint for making artificial intelligence investments profitable.
