Kazakhstan requested $75 million from the World Bank for AI, startups, and future technologies

The decision to approve the project is scheduled for early September 2026, reports Inbusiness.kz. The total cost of the program is $149 million: half is financed by the World Bank, half by the Kazakh side. If the project is approved, implementation will begin in 2027.

The Gist Briefly

• The project of the Ministry of AI and Digital Development of the Republic of Kazakhstan is valued at $149 million, of which $75 million is a World Bank loan. The WB decision is expected in early September 2026, with project closure at the end of 2031.
• The first direction is AI research and infrastructure ($30 million): an AI Center, cloud platforms, robotics laboratories, training programs, and talent attraction measures.
• The second direction is startup development ($39.85 million): support from the idea stage to initial funding, acceleration, international expansion, and a venture fund for investments in ICT, Industry 4.0, and green technologies.
• The third direction is project management and monitoring ($5.2 million): interagency coordination, analytics on gaps in the innovation ecosystem.
• This is already the second major WB loan for the country's digital sector: in 2025, Kazakhstan attracted 42 billion tenge for developing telecommunications infrastructure in 1,123 villages for 2.3 million residents.

Three Tasks of One Project

The structure of the application reflects three fundamentally different levels of the AI ecosystem.

First — infrastructural: the AI Center, cloud platforms, and robotics laboratories are designed to give domestic researchers and developers access to computing resources that are currently unavailable or extremely expensive. Without physical infrastructure — servers, computing power, specialized equipment — AI ambitions remain declarative.

Second — human resources: training programs and talent attraction measures. The FAO calls for deploying AI in service of rural communities — but this requires specialists capable of adapting technologies to local conditions. The shortage of AI personnel in Kazakhstan remains a systemic constraint.

Third — commercial: supporting startups through a venture fund. This is the most market-oriented part of the project — $39.85 million for accelerating and bringing Kazakh technology companies to international markets.

From Villages to the AI Center: The Logic of Two Loans

The two WB loans — last year's and the current one — appear as sequential steps of a single strategy. In 2025, Kazakhstan tackled the task of basic connectivity: internet in 1,123 villages, 500,000 households, eliminating the digital divide between urban and rural areas. Implementation is planned for 2025–2027.

In 2026, the task changes: to connect those already connected to the knowledge economy. The AI Center, startups, venture fund — this is second-level infrastructure operating on top of basic connectivity. The logic is clear: first, a wire to the home; then, tools for creating value through that wire.

Context: Money for an Already Prepared Strategy

The application for the WB loan did not appear in a vacuum. In June 2026, Kazakhstan signed the Pax Silica agreement — a technology alliance of 15 states under the patronage of the USA, focused on AI, critical minerals, and digital infrastructure. Simultaneously, the Ministry of AI and Digital Development is developing a national artificial intelligence strategy.

The WB loan in this logic is not an independent initiative, but a financial instrument for the state architecture that is already taking shape. The three spending directions — infrastructure, personnel, startups — exactly correspond to the three axes of the AI Opportunity Partnership signed with the USA in Washington in June 2026.

Caution: Data Centers and the Water Issue

The AI Center and cloud platforms mean not only server racks, but also colossal demand for electricity and water for cooling. We have already written that New York was the first in the USA to impose a moratorium on the construction of large data centers precisely due to rising electricity bills and the strain on water resources.

Kazakhstan, building its own AI infrastructure under conditions of water stress and energy capacity shortages in the south of the country, must from the very outset embed environmental standards for new facilities — before the problem becomes political, not after.

The Venture Fund as a Key Element

The most non-trivial part of the project is the venture component. A state venture fund focused on ICT, Industry 4.0, and green technologies is an instrument that has worked inconsistently in Kazakhstan. Previous attempts at state venture capital have often faced problems with management and project selection.

This is precisely why a monitoring and analytics block is built into the project structure: $5.2 million for "analytics on gaps in the innovation ecosystem and best practices" — this is essentially a request for institutional honesty: to understand what doesn't work before investing in the same thing again.

Author's Conclusion

$75 million from the World Bank is a serious bid, but not a magic solution. Two consecutive WB loans paint a logical picture: first, basic connectivity; now, the knowledge economy on top of it. But the AI ecosystem needs not only money, but also high-quality project management, transparent startup selection, and genuine attraction of international talent. If the project is implemented with the same systematic approach with which it is formulated on paper, the result could become tangible by 2031.