Ukraine’s Artificial Intelligence Law: How the State Is Preparing Rules Aligned with the EU AI Act
The Ministry of Digital Transformation has begun work on the first specialised law on artificial intelligence. It is planned to be developed during 2026. The document is intended to bring Ukrainian regulation closer to the European AI Act and establish rules for the state, businesses, and technology users. The introduction of such a law means that Ukraine is moving from the stage of recommendations, testing, and voluntary standards to systematic legal regulation. Artificial intelligence is already used not only in commercial products, but also in public services, document processing, citizens’ interaction with services, and the development of digital infrastructure. The more widely technologies enter everyday life, the more important issues of security, responsibility, data protection, and clear rules for developers become.
Time for Action examined why the future law on artificial intelligence may become important for the Ukrainian technology market, public services, and the entry of Ukrainian companies into the international arena. The main purpose of the future law is to align Ukrainian rules with the approaches of the European Union. For Ukrainian developers, this has practical importance. If the requirements for AI products are compatible with the European AI Act, companies will find it easier to work with foreign clients, enter EU markets, and attract partners without having to rebuild their products each time to meet different legal requirements.
For the state, this is also a matter of digital independence and competitiveness. Ukraine seeks not only to use ready-made foreign technologies, but also to create its own solutions, infrastructure, and products that can operate in the public sector, business, education, security, and other areas. The Ministry of Digital Transformation has set an ambitious goal for 2030 to become one of the world’s top three leaders in the level of artificial intelligence implementation in the public sector. This approach involves a transition to the Agentic State model. It refers to a state where digital services not only respond to a person’s requests, but can also offer a needed service in advance or perform basic actions after a voice or text request. This model has already begun to appear in the work of Diia. In the autumn of 2025, a national AI assistant was launched on the portal. During this period, more than 350,000 Ukrainians sent it around 2 million messages. In May 2026, an AI agent appeared in the Diia mobile application. In its first month, 3 million citizens used it to obtain extracts and pay fines. This shows that artificial intelligence in public services is no longer an experiment for a narrow audience. It is becoming a tool for mass use. That is why legislative rules are needed not only for large technology companies, but also for ordinary users.
When an AI agent helps generate a certificate, notifies a person about a fine, or performs another action in a government application, specific questions arise. What data does the system use? How does it explain its decisions? Who is responsible if an algorithm makes a mistake? What protection mechanisms should exist for a person if an automated decision creates a problem? Without clear rules, the development of such services may lose users’ trust. People are ready to use digital tools when they understand that their data are protected and that, in the event of an error, there is an opportunity to receive an explanation and correct the situation. For the state, this means that AI implementation must take place together with responsibility for its operation. Ukraine is beginning to create its own technological foundation for this. Together with NVIDIA, the state is deploying an AI Factory. At the same time, a national large language model, “Siaivo,” is being created with Kyivstar. It is expected to operate on Ukrainian servers and take deeper account of the Ukrainian language, social characteristics, and the needs of public services. A domestic language model may have important significance for state processes. It makes it possible to develop technologies that work better with the Ukrainian language, official documents, local rules, and digital services. It also reduces dependence on exclusively external platforms, where data-use rules, access to tools, or functionality may change without Ukraine having any influence over them.
At the same time, building infrastructure does not guarantee automatic success. A domestic model must be not only technically powerful, but also secure, understandable for users, and suitable for real tasks. That is why the future law will matter not only as a formal European integration requirement, but also as a foundation for the operation of state AI systems. The Ministry of Digital Transformation’s approach to regulation did not begin with strict bans. The state chose a bottom-up model, meaning movement from practice to legislation. First, businesses were given space for testing, and developers received recommendations on how to adapt products to future requirements. Only after that did work begin on legislative boundaries.
Within this approach, a Roadmap, a White Paper on artificial intelligence regulation, and 11 recommendations for the safe use of algorithms in key fields have already appeared. These documents helped companies align with European approaches even before mandatory legislation was introduced. The SandBox regulatory sandbox became a separate tool. In one year, it received 90 applications for testing AI products. Fifteen solutions have already passed an audit involving 23 experts. This format makes it possible to test technologies before their wider market entry, identify risks, and give developers the opportunity to improve a product. For business, this is important because legislation in the AI field can be complex and require significant resources. This is especially relevant for small teams that do not have separate legal departments or large budgets for meeting regulatory requirements. The sandbox makes it possible to understand the state’s expectations before rule violations begin to have legal consequences.
Another step was the introduction of the AI Terms Dictionary 2.0. In 2026, it was made available in a machine-readable format. This matters so that lawyers, developers, government bodies, and companies use the same definitions. In the field of artificial intelligence, terminological confusion can create problems no less serious than the absence of the law itself. Different understandings of what an algorithm, an automated decision, a model, a high-risk system, or developer responsibility mean complicate the work of the market. A common language between technical specialists and lawyers is necessary so that rules can not only be written, but also applied in practice.
Ukraine is also trying to unite the market around the ethical use of AI. Fourteen leading companies, including Grammarly, MacPaw, SoftServe, and Uklon, have created the country’s first specialised organisation. Its emergence shows that business is interested not only in freedom to develop technology, but also in common rules that can increase trust in Ukrainian products. At the international level, Ukraine has signed the Bletchley Declaration and the Council of Europe Framework Convention on Artificial Intelligence. HUDERIA workshops dedicated to assessing the impact of algorithms on human rights have also been held. This means that the future law must take into account not only the interests of the state and business, but also the risks that arise for people when automated systems are used. The main task of the new law is to find a balance. Excessively strict requirements may slow the development of Ukrainian companies, especially those that are only creating their first products or working in small teams. Rules that are too weak may leave users without adequate protection when their data and rights can be affected by algorithmic errors or non-transparent decisions.
Ukraine already has examples of AI working with mass public services. Therefore, the future law is needed not for abstract regulation of technologies, but for specific situations: obtaining documents, working with data, automated notifications, electronic services, digital interaction with the state, and the development of business products. The specialised law on artificial intelligence may become the point where Ukraine’s technological ambitions receive a legal foundation. It is intended to help the state implement AI without losing control over security, help businesses enter international markets, and help citizens use digital services with an understanding of how their rights are protected.
How precise and balanced these rules will be will determine not only the pace of development of Ukraine’s AI market, but also the level of trust in technologies that are increasingly influencing everyday life.
About the Author
Related posts:
OpenAI Moves Into Health: How Artificial Intelligence Is Changing Digital Medicine 
Artificial Intelligence in Ukraine: Why It Is Not a Must-Have Today but Will Shape Careers Tomorrow 
Artificial Intelligence at DTEK: How Big Business Implements AI in Practice 
Artificial Intelligence in Government: How Diia.AI Works and Why It’s More Than Just a Chatbot
