Russia and Belarus team up on an AI system "based on traditional values."
The two countries plan to develop a sovereign AI they say can be "trusted, relied upon," and gives "objective information," according to Sergei Glazyev, secretary of their Union State.
Glazyev flagged concerns about foreign AI models, especially from the U.S. and China, claiming they manipulate younger users. 
“Recent deep testing of Western-type artificial intelligence revealed that one prominent chatbot had demonstrated racist and extremist tendencies, including the glorification of fascism,” Sergei Glazyev stated, without naming the chatbot or providing proof.
The goal is to back entrepreneurship and innovation, but "preserving fundamental and traditional values." Glazyev didn’t specify what those values are but warned entrepreneurship without them is “destructive, senseless and harmful.”
This push fits Moscow and Minsk’s drive for digital independence from the West, boosted by Ukraine war sanctions.
But independent research raises red flags. A Ghent University study found Russian models YandexGPT and GigaChat top the list in political censorship among 14 global AI systems, including Chinese ones.
These Russian models dodge sensitive political questions, often telling users they can’t discuss certain topics or advising them to look elsewhere.
Russia and Belarus are building an AI that fits their own worldview — filtered, controlled, and tied to “traditional values.” How that will affect transparency or trust remains to be seen.
