Abstract

Kazakhstan has recently pursued a state-led push towards using AI technologies in economic modernization and administrative efficiency. Initiatives such as the Alem.AI center and the introduction of an AI advisory system at Samruk-Kazyna signal an unusually top-down approach, in contrast to global patterns, in which AI development is typically driven by the private sector. This study examines how these developments are represented in the national news cycle and asks whether AI coverage in Kazakhstan reflects global narratives or exhibits distinct domestic patterns.

We conducted a dictionary-based analysis of more than 600 news articles published between January 2023 and December 2025 across four major outlets: the state-aligned Informburo.kz, the pro-government commercial Tengrinews, and the independent outlets Vlast.kz and Ulys Media. Using bilingual (Russian/English) keyword dictionaries, compound frame indices were constructed to show how frequently the themes of “modernization”, “control”, and others appear in news articles. The expectation is that state-aligned media frames AI as “modernization,” and independent outlets focus on “ethics” and “risks” more.

Our analysis reveals a dominant trend across all news outlets. AI is overwhelmingly framed in terms of modernization, efficiency, and innovation, while the coverage of ethics, privacy, or transparency appears only rarely. Approximately 90% of articles contain no references to ethics, privacy, or transparency. Contrary to expectations, independent outlets do not provide more critical coverage: they publish fewer AI-related articles overall and do not engage more extensively with ethics, privacy, or transparency issues than state-aligned media.

These patterns suggest that AI functions as a narrative tool for reinforcing technological optimism in Kazakhstan, contributing to a public discourse centered on efficiency rather than accountability. This study contributes to the existing literature on soft authoritarianism by showing how AI narratives narrow public debate on ethics through specific framing rather than censorship. It also challenges assumptions about media ownership, demonstrating that editorial independence does not necessarily translate into more critical coverage of emerging technologies. Finally, the paper also extends research on the AI governance literature to the understudied Central Asian context, providing empirical evidence on how political and media systems shape technological discourse.