On 18 January 2026, Law of the Republic of Kazakhstan “On Artificial Intelligence,” adopted on 17 November 2025, enters into force. The legislation introduces, for the first time, a comprehensive legal framework governing the status of artificial intelligence (AI) systems, their use, and liability for the outcomes they produce.
The law aims to establish a foundation for the safe, transparent, and responsible deployment of AI technologies, promote innovation, and protect the rights of participants in the digital economy. Prior to its adoption, these matters were not specifically regulated, creating legal uncertainty for businesses and rights holders.
Under the law, artificial intelligence is defined as a functional capability to imitate human cognitive functions and generate results comparable to, or exceeding, those of human intellectual activity. Images, videos, audio materials, texts, or combinations thereof created or modified by AI systems that simulate a person’s appearance, voice, behavior, or events that did not actually occur are classified as synthetic AI outputs.
The authorized governmental body in the field of AI is the Ministry of Artificial Intelligence and Digital Development of the Republic of Kazakhstan.
The law establishes key principles for AI use, including transparency, safety, respect for third-party rights, and explainability of AI-driven decisions. Liability for the actions or outcomes of AI systems rests with the developer, owner, or user of the system, prompting companies to implement internal policies and contractual safeguards to mitigate legal risks.
Special attention is given to copyright issues related to AI-generated content. In particular, the law provides that:
- copyright protection applies only where a human has made a creative contribution;
- creatively formulated text prompts are recognized as copyrightable works;
- the use of works for training AI models does not qualify as free use for educational or scientific purposes;
- AI training does not grant the right to reproduce, adapt, display, perform, or distribute works without the consent of the rights holder;
- works may be used only in the absence of a prohibition expressed in machine-readable form.
For businesses, the new rules underscore the importance of defining rights to AI-generated and AI-assisted content in advance, documenting employees’ creative contributions, and ensuring compliance with third-party rights when training and deploying AI systems.