YouTube Algorithms Prioritise Russian Content Over Kyrgyz in Central Asia
A collaborative study by US and Kyrgyzstan universities finds that searches for children's content in Kyrgyz yield predominantly Russian results, raising concerns about language preservation.
Researchers from the United States and Kyrgyzstan have released findings indicating that YouTube's search and recommendation algorithms systematically prioritise Russian-language content over Kyrgyz, an indigenous language. This bias is particularly pronounced when users search for children's topics such as cartoons, fairy tales, and mermaids. Simulated user behaviour experiments demonstrated that even after viewing Kyrgyz-language videos, the algorithm continued to recommend Russian content, effectively constructing Kyrgyz children as an audience for Russian media.
The study was conducted by a team of six researchers across four universities, including the University of Michigan and UC Berkeley. The group simulated user behaviour on the platform and collected nearly 11,000 unique search results and video recommendations. The findings show that Kyrgyz-language searches for popular kid interests often did not yield content in Kyrgyz. Even after watching 10 children's videos featuring Kyrgyz speech to demonstrate a strong desire for it, the simulated users received fewer Kyrgyz-language recommendations for what to watch next than bots showing no language preference at all.
Despite the existence of significant Kyrgyz-language content on the platform, the algorithm fails to promote it to Kyrgyz-speaking children. For instance, the D Billions channel, which is the 35th-most viewed globally and has nearly one million subscribers, remains largely invisible in these specific algorithmic pathways. The research indicates that only 2.7 per cent of analysed videos appearing in Kyrgyz searches feature ethnically Kyrgyz people. This dynamic suggests that YouTube socialises youth to view Russian as the default language of entertainment and technology, while viewing Kyrgyz as uninteresting.
The issue is set against a complex historical backdrop where Kyrgyzstan was under Russian control for a century until 1991. While Kyrgyz survived as a widely spoken language among adults, Russian remains crucial for commerce and success. Under Soviet rule, learning Russian was mandatory for advancement, a cultural legacy that persists today. Although schools require children to learn at least some Kyrgyz, many spend several hours daily online, with YouTube being the leading activity. Anthropological researcher Ashley McDermott has long observed that rural children in Kyrgyz-dominated villages are spontaneously learning Russian, often blaming YouTube for this shift.
YouTube has publicly committed to amplifying indigenous voices but has not responded to specific requests for comment regarding this issue. The researchers are currently attempting to meet with YouTube's parental controls team to discuss implementing language filters to mitigate the bias. The team argues that their work is the latest to show how online platforms can reinforce colonial culture and influence offline behaviour, noting that the platform effectively constructs Kyrgyz children as audiences for Russian content.
While the study relies on simulated user behaviour rather than direct observation of real-time user data, the researchers argue the simulation accurately reflects algorithmic logic. They suggest potential remedies to parents, such as creating playlists of Kyrgyz-language content or sharing devices with their children. When the researchers simulated adult users watching non-kid's content in Kyrgyz, they received predominantly Kyrgyz-language recommendations, highlighting the specific nature of the bias affecting younger users.
