Kyivstar, Ukraine’s largest telecommunications operator, has signed a memorandum of cooperation with the Ministry of Digital Transformation to develop a Ukrainian language model (LLM). This initiative aims to ensure data sovereignty by keeping data on local infrastructure, thus avoiding concerns over storing data abroad. The project, expected to take nine months, will use open-source solutions and will not rely on public funding, with a beta version set for release by December 2025.
Deputy Prime Minister Mykhailo Fedorov highlighted that this LLM will enable the creation of AI services for citizens and enhance internal processes in various sectors, including education and regulatory analysis. The model will be uniquely tuned to understand Ukrainian dialects and cultural context, providing more accurate responses than global counterparts. Kyivstar will lead the operational aspects, establishing a project office and covering finances, while a committee will oversee data handling and model training.
Source link
