Quick overview

In 2026 the Ministry of Digital Transformation together with Kyivstar plan to launch a national large language model (LLM) that will serve as a platform for solutions in public administration, defense, business, science and education. Alongside the model they will create AI Factory — a state infrastructure for training and the safe operational use of AI based on Nvidia equipment. The plans were announced by Danylo Tsvok in an interview with Channel 24.

"We are working on a systemic infrastructure: this is not only the model, but also a production environment for the safe deployment of AI solutions in public services and critical infrastructure."

— Danylo Tsvok, head of WINWIN AI Center of Excellence (interview with Channel 24)

What this means for security and sovereignty

Having its own LLM reduces dependence on foreign cloud services and closed models, which directly affects the security of state data and military applications. Essentially, this is an investment in digital sovereignty: from control over model configurations to the ability to perform local training of critical systems.

AI Factory: infrastructure or challenge?

AI Factory is presented as a complex of equipment for model training, data storage and access control. Using Nvidia solutions provides a technical advantage, but at the same time raises questions about energy supply, cybersecurity and a secure supply chain. Success will depend on the operational model, security standards and auditing.

Economy, science and international confidence

The project could become a catalyst for Ukrainian startups and research teams: access to local models and infrastructure lowers barriers to entry. It is also a signal to partners: Ukraine's ambition to enter the top three in AI adoption by 2030 is accompanied by concrete steps — updating the national strategy, a recent rise in global rankings (+14 positions) and Diia.AI's victory in a European competition for digital public services.

Risks and what needs to be controlled

Even a patriotic initiative requires strict safeguards: transparent data governance, independent model audits, mechanisms to combat bias and human oversight in critical decisions (especially in the judiciary). Integrating AI into courts could increase efficiency, but only with clear rules of use and the right to appeal decisions.

Conclusion

The launch of a national LLM and AI Factory is more than a technological project; it is a geopolitical and economic gesture. It is now important that declarations turn into operational discipline: transparent procedures, international cooperation and protected infrastructure. Whether Ukraine can use this opportunity to genuinely enhance security and competitiveness depends on how quickly and responsibly the plans are implemented.