Standards of Trust and Safety: KazInSt Discussed Draft National AI Standards
On August 12, 2026, at 10:00 AM, the RSE "Kazakhstan Institute of Standardization and Metrology" (KazInSt) hosted a technical discussion on key draft national standards in the field of Artificial Intelligence. Organized online via the ZOOM platform under the 2026 State Standardization Plan, the event brought together leading industry experts, software developers, academic researchers, and representatives of government agencies.
The primary objective of the meeting was to conduct a comprehensive technical review of the proposed regulatory drafts, evaluate their alignment with the current legislation of the Republic of Kazakhstan, and ensure harmonization with international best practices. Participants thoroughly analyzed the core content of the documents and examined practical mechanisms for integrating these standards into the national digital ecosystem.
The expert panel evaluated the following key draft standards:
ST RK ISO/IEC TS 6254 "Artificial Intelligence 鈥� Explainability of AI Systems": This document defines concepts and methodologies for ensuring the transparency and explainability of AI algorithms. For developers and tech companies, the standard offers a structured toolkit to improve system reliability, predictability, and safety. For end-users and regulatory bodies, it establishes a methodological foundation to assess trust and identify potential algorithmic bias or unexpected behavior.
ST RK ISO/IEC 42005 "Artificial Intelligence 鈥� AI System Impact Assessment": This standard provides comprehensive guidelines and practical recommendations for evaluating the societal, ethical, and individual impacts of AI technologies. It sets clear protocols for conducting impact assessments across all stages of the AI lifecycle 鈥� from initial model design and training to deployment and decommissioning.
During the interactive session, participants exchanged professional insights and submitted constructive feedback and detailed proposals to refine the draft texts.
The adoption of these standards represents a major milestone in building a trustworthy, safe, and transparent regulatory framework for artificial intelligence in Kazakhstan. The active engagement of experts in such technical discussions elevates the quality of national standards, strengthens the country's scientific and methodological foundation, and supports quality infrastructure across standardization, certification, and metrology. All feedback collected during the meeting will be compiled and incorporated into the final versions of the documents.
