The Deputy Prime Minister — Minister of Artificial Intelligence and Digital Development, Zhaslan Madiyev has approved the rules for the formation and publication of lists of trusted high-risk artificial intelligence (AI) systems. The document comes into effect on 26 April 2026.
WHAT THE ORDER PROVIDES
The Order dated 10 April 2026 establishes the procedure for sectoral state bodies to compile lists of trusted high-risk AI systems.
Such lists are created to increase trust in the technologies and to disseminate best practices for the use of AI in various fields. They are published on the internet resources of state bodies.
HOW THE LISTS ARE FORMED
Inclusion of an AI system in the list is carried out based on an application from the owner or possessor.
This requires the submission of:
- an application;
- a certificate of entry of information into the State Register of Rights to objects protected by copyright, or a copyright agreement;
- an audit report.
If the package of documents complies, the sectoral state body includes the system in the list.
DEADLINES AND PUBLICATION PROCEDURE
After the decision is made, information about the AI system is placed on the official internet resource of the state body within 5 working days.
The lists are published in open access on special pages of websites intended for regulatory and reference materials. Users are provided with the ability to view and, where available, download the documents.
HOW THE LISTS ARE UPDATED
When the list is updated, its revised version is published on the website. No other notification mechanisms or clarifying procedures are specified in the materials.
SIGNIFICANCE OF THE DOCUMENT
The adopted rules establish a unified mechanism for selecting and publicly disclosing information about high-risk AI systems. This may contribute to increasing the transparency of their use in state and sectoral processes.