"AI Factory", 30,000 specialists and 500 times more powerful calculations: what Ukraine plans to do with artificial intelligence by 2030

Why do they do this at all?

The authors of the Strategy state that despite developed e-governance and strong IT human resources potential, Ukraine lags behind in computing infrastructure, access to data, financing of developments, and use of AI by businesses. Only 4.6 % of business entities reported using at least one AI function in 2025 — 20-30% less than the EU and OECD average.

The state wants to correct the situation in three stages: first, accelerate the security and defense sector as a catalyst (there is already practical experience from there), then scale the solution to public administration and key sectors of the economy, and in the third stage, move on to the export of intellectual services and full integration into the European digital market.

What awaits IT professionals and businesses

Money and benefits. Grant instruments and acceleration programs for AI startups are planned through the Innovation Development Fund, co-financing with private venture funds (matching grants), as well as targeted grants for higher education in AI specialties.

«Artificial Intelligence Factory.» By the end of 2028, a hardware-software complex for developing, training, and deploying models should be operational — essentially, a shared infrastructure that government agencies, SMEs, and researchers can use so they don’t have to pay for expensive computing resources themselves.

Ukrainian Language Model. The state wants to ensure the emergence of at least one large language model (LLM) optimized for the Ukrainian language and published under an open license by 2026, with the possibility of commercial use.

«Sandbox for AI and Blockchain Solutions.» A regulatory sandbox project where companies can test solutions before full legislation is passed. The goal is 200 tested projects by 2030.

Human resources. Currently, there are about 6,100 specialists working in the field of AI and machine learning (data from 2023), mainly in the field of computer vision. The goal for 2030 is 30,000 specialists, 50% of whom are women (currently 17%). In parallel, mass retraining programs are being prepared: at least 1 million people should receive confirmed AI skills by 2030.

New requirements in procurement. The government plans to approve standard technical requirements for software and information systems with AI, as well as introduce interoperability requirements in public procurement — that is, suppliers will have to adapt to new standards.

The target is that 75% of companies with more than 10 employees should use AI as a standard tool by 2030 (compared to 5% now).

What awaits the defense sector?

Defense is one of the main priorities of the Strategy, and it is here that the most concrete steps are planned:

«Defense AI Factory» is a platform for coordinating developments between the state, the Armed Forces, and private companies, together with the Defense Data Coordination Center.

Military training. Military AI literacy programs for military personnel, training at least 2,000 defense AI specialists by 2030 through boot camps and rotation between the Ministry of Defense, business, and academic institutions.

CAIO in the power unit. Key agencies in the security sector will have AI chiefs at the deputy chief level, with authority over infrastructure, data protection, and personnel.

Secure data infrastructure. They are planning a «Military Data Lake» — a secure platform for aggregation of military data, as well as a closed repository of code and models for verified participants in the MilTech ecosystem (Sovereign AI Forge).

Multi-agent systems. By the end of 2028, a unified analytical multi-agent system for the Ministry of Defense and the General Staff and a unified decision-making support system for the command staff of the Defense Forces should be operational.

Faster implementation cycle. The time from confirming the operational need to deploy an AI solution on the ground should be reduced from the current undefined figure to less than 8 weeks (interim target — 12 weeks).

Export control. A more flexible licensing regime is promised so that defense developments can be commercialized in civilian and allied markets.

What awaits civil servants and citizens?

«Agent State». By 2030, at least 100 administrative services on the Unified Portal of Electronic Services should be provided using AI (currently 1). In parallel, internal processes of government bodies are being automated.

AI Transparency Panel. A section will appear on the Ministry of Digital Affairs website where information about algorithms used in public services will be published.

Training of civil servants. At least 500 civil servants of categories «A» and «B» will undergo a specialized program on AI project management.

Regulation is phased in. Until 2028, mainly «soft law» instruments will be in effect — voluntary codes of conduct, sandboxes, procurement standards. After 2028, a profile law on AI and a Single Digital Regulator will come into effect, which will perform market surveillance functions in accordance with the EU Regulation on AI.

Education and healthcare

Schools and universities will update state standards to take into account AI skills, launch a digital educational platform with a module to support individual educational trajectories and the «Mriia-first» architecture on the «Diya.Osvita» portal. The goal is 100% of education seekers with access to AI-based learning tools by 2030 (currently no data), and 80% of teachers who have undergone appropriate training.

In healthcare, where over 5 billion medical records on 34 million patients have already been accumulated, a roadmap for the ethical use of AI is being planned, as well as competence centers based on medical education institutions and training of healthcare workers — the goal by 2030: 700 healthcare entities using AI.

The main challenges that the government itself recognizes

The authors of the document do not hide the weaknesses:

• the computing power of the state infrastructure for AI is only 0.1 petaflops — the target for 2030 is 50 petaflops, i.e. a 500-fold increase;
• R&D spending in Ukraine is about 0,3% of GDP compared to 2,3% in the EU;
• the venture capital market remains limited (about UAH 22 billion in 2021);
• energy infrastructure is regularly attacked, which complicates the construction of new data centers;
• legal uncertainty regarding the use of data (including copyrighted data) to train AI models.

The implementation was divided into two regulatory stages: 2026–2028 — «soft law» and capacity building, 2028–2030 — transition to full-fledged legislative regulation and the work of the Single Digital Regulator. Financing — within the limits of expenditures of profile managers of budget funds and other sources not prohibited by law, i.e. the document does not provide for a separate special budget for the Strategy.

Monitoring of implementation is entrusted to the Ministry of Digital Affairs, which will prepare reports annually taking into account the established indicators.

Your photos and voice queries can train Google’s AI. How to check your settings

On the topic

Your photos and voice queries can train Google’s AI. How to check your settings

Over 60 million tokens without drawdown. De Novo company shared the results of testing Ukrainian LLM

On the topic

Over 60 million tokens without drawdown. De Novo company shared the results of testing Ukrainian LLM

Read the country’s main IT news in our Telegram

On the topic

Read the country’s main IT news in our Telegram