Kazakhstan introduces Central Asia’s first comprehensive AI Regulation

Kazakhstan has adopted a groundbreaking national law on artificial intelligence, making it the first country in Central Asia to introduce a comprehensive AI specific regulatory framework. The Law on Artificial Intelligence entered into force on January 18, 2026, and is designed to foster innovation while ensuring safety, transparency, accountability, and a human centric approach to AI development and use.

The new law does not rewrite Kazakhstan’s existing copyright system, which continues to be based on the Civil Code, the Law on Copyright and Related Rights, and international agreements such as the Berne Convention. Instead, it supplements these rules with targeted provisions on AI generated content and data use.

Copyright for AI generated works

Under the law, AI generated works are eligible for copyright protection only if a human makes a creative contribution to their creation. Purely autonomous outputs produced by AI systems, without significant human intellectual or creative input, are not protected, as authorship under Kazakh law remains reserved for natural people.

Textual prompts that reflect human intellectual or creative effort are recognized as potential copyright protected objects in line with general copyright principles. This means that well-crafted, original prompts may themselves be treated as protected expressions if they meet the usual criteria for originality.

AI training and opt-out for data use

Training AI models on copyrighted works is generally allowed and is not automatically considered an infringement of exclusive rights such as reproduction or adaptation. However, authors or rights holders can block such use by clearly indicating their objection in a machine readable format, for example, via metadata or opt-out signals. This establishes an opt-out model for data use in AI training, shifting the burden of compliance onto AI developers and operators.

Transparency, liability, and prohibited uses

The law also introduces several related requirements:

• AI generated or synthetic content, including images, video, audio, and text, must be labeled or marked in both visual and machine readable forms to ensure transparency and to help prevent misuse such as deepfakes.
• Liability for AI related damages is governed by the Civil Code, with responsibility typically assigned to developers, owners, or operators of the AI system.
• AI systems that violate rights, such as those that infringe copyright or generate prohibited content, are expressly prohibited.

Global context and practical implications

Kazakhstan’s approach reflects broader international trends, particularly the emphasis on human involvement as a condition for copyright protection, which echoes certain interpretations in the United States and the European Union. At the same time, the law is tailored to promote ethical AI development while safeguarding creators’ interests.

Some questions remain open, including the precise threshold for what counts as a “creative contribution” and how machine readable opt-out mechanisms will be enforced in practice. These issues are likely to be clarified through court decisions, administrative guidance, or future amendments.

For creators and businesses using AI in Kazakhstan, documenting human input, such as detailed prompts, editorial decisions, or creative edits, has become especially important for supporting future copyright claims or defenses.

The attorneys at the IPR Group are already advising clients on how to navigate the new legal landscape, helping them address potential disputes, licensing issues, and compliance challenges arising from the AI law and related changes in Kazakhstan and other CIS jurisdictions.