Ukraine’s Ministry of Digital Transformation has inked a deal with Kyivstar to help develop the country’s first national large language model (LLM) which will be trained entirely on Ukrainian-language data.
The model will incorporate regional dialects, cultural context and national terminology, aiming to deliver AI services tailored to local needs. The LLM aims to support AI use cases in sectors such as government, healthcare, finance and education, claiming to offer accurate, locally relevant insights that are not currently addressed by general-purpose global models.
The model will be built on open-source architectures and trained using Ukrainian data sources with a focus on data sovereignty and national security. All data will be stored and processed within the country.
Development will be led by Kyivstar in collaboration with the Ministry and the WinWin AI Centre of Excellence, Ukraine’s national platform for technology testing. The first version of the model is expected by December 2025.
The project forms part of the investment from Kyivstar’s parent company Veon to send $1 billion on rebuilding Ukraine’s digital infrastructure between 2023 and 2027.
Veon CEO Kaan Terzioglu hailed the agreement as “a major milestone” in building a secure and culturally relevant AI ecosystem for the country. “Ukrainian LLM will empower users to access augmented intelligence tools with the full cultural context and depth of their native language and national resources”.
Source: Mobile World Live
Image Credit: Stock Image