Yakutia's AI Is to Speak the Yakut Language
Neural networks capable of translating official documents into the Yakut language are expected to begin pilot operations this year. If the project succeeds, the technology could provide a model for other regions where national languages are used alongside Russian.
Yakutia is preparing to introduce AI technologies across the public sector. Neural networks will automatically translate official documents and draft legislation into the Yakut language, while also processing citizen requests submitted in Yakut. The initiative is a joint project of the Yakutia AI Laboratory, the Republican Center for Infocommunication Technologies (RCIT), and the North-Eastern Federal University. A pilot launch is scheduled for 2026. The system builds on speech recognition and speech synthesis technologies developed by local researchers through an earlier grant-funded project.
For the republic, the initiative marks an important step in bringing one of its official languages into the digital era. It also advances AI development in one of the field's most technically demanding areas – processing Russia's national languages. That work requires new AI models, which are already under development at the North-Eastern Federal University. Once deployed, the technology will allow residents to submit requests, obtain official information, and work with government documents in their native language.
The Future: Yakut-Language Chatbots
Over time, large language models could be fully integrated into government services to assist residents. They could automatically translate laws and secondary legislation, handle voice-based citizen requests, transcribe spoken language, and generate spoken versions of official information. The same technology could also support Yakut-language chatbots for public services, education, culture, and social assistance.
The methodology and technical expertise developed through this project could later be adopted by other Russian regions with national languages. Beyond Russia, countries that recognize more than one official language may also find the approach useful.
From a Translation Tool to an AI Ecosystem
The effort to bring the Yakut language into digital assistants began several years ago. In 2020, Yakut was added to Yandex Translate. As generative AI became more widely available, work expanded to integrate national languages into AI systems as well. In 2024, Yakutia launched a campaign to collect voice recordings for AI training in the Yakut language. That same year, the Yakutia AI Laboratory was established to address socially significant challenges across multiple sectors. Among its projects, the laboratory also supported machine text recognition technologies.
In 2025, the North-Eastern Federal University developed automatic translation systems for Russian-to-Yakut, Yakut-to-Russian, and Yakut-to-English translation. Development of a virtual assistant in the Yakut language brought together the North-Eastern Federal University, the Yakutia AI Laboratory, the National Library, and the Sever (North) Research and Education Center. Then, in 2026, the Yakut language was added to GigaChat, the AI assistant developed by Sber.
Public Services Without Language Barriers
Beyond helping preserve the national language, Yakut-language AI could make government services more accessible to residents. If the pilot proves successful, Yakutia could become one of Russia's first regions to systematically integrate a national language into AI-powered public services. In practice, the experience gained there could later be extended to other languages spoken by Russia's indigenous peoples.
The project also carries an important challenge. Much will depend on the accuracy of AI translation for legal and administrative documents. Translation errors in citizen requests could have significant consequences. For that reason, specialists will be required to review the system's output during the initial stages of deployment.
