Finland’s Ministry for Foreign Affairs has introduced a secure artificial intelligence assistant designed to process classified government information, marking a new step in the country’s efforts to bring AI into sensitive public administration work.
The system, called UMAI, entered pilot use on 13 October 2026. It operates in a secure environment approved to handle information classified at Finnish national security level IV, known as TL IV, according to an announcement from the ministry on Tuesday.
TL IV is the lowest of Finland’s national security classification levels. However, information in this category still requires protection and must be handled according to specific security procedures. The approval allows UMAI to be tested in a government environment where confidentiality requirements limit the use of conventional AI services.
The ministry has initially restricted access to a small group of users who will test the assistant and assess how well it meets the needs of government work. The pilot is intended to establish how artificial intelligence can support staff while keeping control of information processing, technical infrastructure and security procedures within the government system.
The initiative also involves the Government ICT Centre Valtori, with both organisations participating as pilot users in the development of a shared AI platform for Finland’s central government. The aim is to build experience that can inform the platform’s development and help establish common requirements for secure AI services across public administration.
Feedback from the initial users will help authorities identify technical needs, improve the system and assess whether similar services could be introduced in other government departments. The project is also intended to strengthen Finland’s expertise in artificial intelligence and develop consistent practices for its use across public institutions.
The introduction of UMAI comes as governments face the challenge of using AI tools without exposing sensitive information to systems that do not meet their security requirements. By testing the assistant within an approved environment, the ministry is exploring how public officials can use AI while maintaining government oversight of the technology and the information it processes.
Foreign Minister Elina Valtonen connected the initiative to Europe’s broader efforts to strengthen its control over critical technologies and reduce dependence on external providers.
“In the age of AI, Europe must retain its ability to make its own choices and build sufficient capacity, competitiveness and capability to control critical technologies,” Valtonen said in the ministry’s official statement.
She also stressed the need for European countries to develop the expertise and resources required to assess AI systems and decide how to use them according to their own needs.
The project aligns with the principle of European digital sovereignty, which centres on maintaining control over data, digital infrastructure and critical technologies. For Finland’s public administration, developing secure AI capabilities could provide a way to explore the technology in areas where confidentiality and information security are essential.
UMAI remains in the pilot stage, and its future development will depend on the experience gathered during testing and the requirements identified by participating organisations. The initial rollout does not establish how widely the system will be adopted, but it gives Finnish authorities an opportunity to assess the practical requirements of using AI to handle protected government information.
