RESOLUTION OF THE CABINET OF MINISTERS OF THE KYRGYZ REPUBLIC

of December 2, 2025 No. 770

About approval of requirements to technique of assessment of danger of systems of artificial intelligence, and also requirements for systems of artificial intelligence of the increased danger

According to Articles 193, 194 Digital codes of the Kyrgyz Republic, articles 13, of the 17th constitutional Law of the Kyrgyz Republic "About the Cabinet of Ministers of the Kyrgyz Republic" the Cabinet of Ministers of the Kyrgyz Republic decides:

1. Approve:

1) Requirements to technique of assessment of danger of systems of artificial intelligence according to appendix 1;

2) Requirements to risk management concerning systems of artificial intelligence of the increased danger according to appendix 2;

3) Requirements to the characteristics of systems of artificial intelligence of the increased danger providing openness, explanation, submission to control, accuracy, reliability and digital stability of such systems according to appendix 3;

4) Quality requirements of digital data for systems of artificial intelligence of the increased danger according to appendix 4;

5) Requirements to technical documentation on systems of artificial intelligence of the increased danger according to appendix 5.

2. To the owners of systems of artificial intelligence applied in the Kyrgyz Republic irrespective of forms of business, departmental (industry) accessory and patterns of ownership:

1) to take the measures following from this resolution;

2) to provide evaluating danger of systems of artificial intelligence at the corresponding stages of their lifecycle by the techniques developed and approved independently according to the requirements approved by this resolution;

3) in case of reference by results of assessment of danger of systems of artificial intelligence to systems of artificial intelligence of the increased danger:

- provide their compliance to the mandatory requirements approved by this resolution;

- implement and for all lifecycle to support risk management system;

- develop and post on the websites technical documentation for such systems and provide general availability of other required documents according to the procedure, established by the Digital code of the Kyrgyz Republic and this resolution.

3. To impose control of execution of this resolution on management of control of execution of decisions of the President and the Presidential Administration Cabinet of Ministers of the Kyrgyz Republic.

4. This resolution becomes effective after fifteen days from the date of entry into force of the Digital code of the Kyrgyz Republic.

Appendix 1

to the Resolution of the Cabinet of Ministers of the Kyrgyz Republic of December 2, 2025 No. 770

Requirements to technique of assessment of danger of systems of artificial intelligence

Chapter 1. General provisions

1. Requirements to technique of assessment of danger of systems of artificial intelligence (further - Requirements) are developed according to the Digital code of the Kyrgyz Republic and establish procedure for determination and application of technique of assessment of danger of systems of artificial intelligence by their owners.

2. The terms and concepts used in these Requirements are used in the values specified in the Digital code of the Kyrgyz Republic.

3. All systems of artificial intelligence applied in the Kyrgyz Republic are subject to assessment of danger which their application can create for the protected benefits which treat:

1) life and human health;

2) rights and freedoms of man and citizen;

3) environment;

4) defense capability of the state;

5) homeland security of the state;

6) public order.

4. The purposes of evaluating danger of systems of artificial intelligence are:

1) identification of the bases allowing to carry the evaluated systems of artificial intelligence to systems of artificial intelligence of the increased danger, falling under the mandatory requirements established by the Digital code of the Kyrgyz Republic;

2) damnification risk identification to the protected benefits which can result from use of the evaluated systems of artificial intelligence;

3) assessment of probability of approach and the amount of damnification to the protected benefits which can result from use of the evaluated systems of artificial intelligence.

5. Results of assessment of danger are the decisions on availability or lack of the bases for reference of the evaluated systems made by owners of systems of artificial intelligence to systems of artificial intelligence of the increased danger.

6. Assessment of danger is performed by owners of systems of artificial intelligence:

1) at design stage of systems of artificial intelligence or planning of changes in them;

2) upon completion of development and prior to use of systems of artificial intelligence or the planned changes made to them;

Disclaimer! This text was translated by AI translator and is not a valid juridical document. No warranty. No claim. More info