Kazakhstan’s Parliament passes Artificial Intelligence law
Tengrinews.kz — Members of Kazakhstan’s Mazhilis (Lower House) have approved the Senate’s amendments to the law on Artificial Intelligence (AI) and related legislation, officially passing the bill and sending it to the President for signature.
Last week, the Senate returned the draft law to the Mazhilis for revision, proposing several key changes — including the introduction of mandatory liability insurance for potential harm caused by AI systems. Previously, the law provided only for voluntary insurance.
Another Senate amendment now requires immediate preventive measures to avoid harm or eliminate risks, rather than simply minimizing damage once identified.
The Mazhilis agreed to all proposed amendments. During today’s plenary session, the Committee on Economic Reform and Regional Development recommended adopting the revised version, which was then approved by the deputies.
With this vote, the law on AI has been formally adopted and will now be sent to the President for final approval.
Background
In Kazakhstan’s legislative process, the Mazhilis passes bills that are then reviewed by the Senate. The Senate may either approve the bill and forward it to the President or return it for reconsideration — as it did with the AI law.
Kazakhstan joins a growing list of countries developing legal frameworks for artificial intelligence:
The European Union pioneered a comprehensive AI Act, setting obligations, penalties, and safeguards for high-risk systems.
South Korea passed a framework AI law, while China, the U.S., U.K., and UAE regulate AI through specific directives.
Japan, Brazil, and India are still drafting similar legislation.
What the Kazakh AI law includes
Adopted in September, the Kazakh law establishes human rights, transparency, and national values as guiding principles. It prohibits emotion analysis, restricts the transfer of confidential data to neural networks, and introduces mandatory labeling of deepfakes to prevent misinformation.