Highlights

  • Kyivstar to lead development of Ukraine’s national LLM using Google Gemma.
  • Project aims to secure and refine Ukrainian linguistic, cultural, and contextual data.
  • VEON sees the initiative as part of a broader effort to enhance local AI ecosystems.

VEON Ltd. (Nasdaq:VEON) announced that Kyivstar (Nasdaq:KYIV; KYIVW), working with the WINWIN AI Center of Excellence under Ukraine’s Ministry of Digital Transformation, has selected Google Gemma as the foundation for the country’s national large language model. The effort will use Google’s Vertex AI infrastructure to provide the required computing capabilities.

Kyivstar will take a central operational role in the development. The model aims to capture Ukrainian dialects, terminology, and historical context. A major focus is secure handling of national datasets, especially for sectors such as government, healthcare, and financial services.

Selection of Google Gemma

Google Gemma was chosen following extensive evaluations. The selection reinforces technology ties between Ukraine and the United States. Kyivstar’s Nasdaq listing in August 2025 also contributed to increased cooperation. By choosing an open-source model, Ukraine aims to create an LLM aligned with national language needs and security standards.

“We are building the Ukrainian LLM on a ready-made open-source model. The main task in development is to train it on our unique data further. When choosing a model, we focused on how well it already handles Ukrainian-language texts and how controllable it is during additional training. This will help minimize linguistic and ethical risks in our LLM,” said Danylo Tsvok, Chief AI Officer at the Ministry of Digital Transformation and CEO of the WINWIN AI Center of Excellence.

Strategic Importance for Ukraine’s Digital Sector

According to VEON Group CEO Kaan Terzioglu, “Kyivstar and the Ukrainian Digital Ministry have taken a major step forward today. With a sovereign Ukrainian LLM, Ukrainian consumers, businesses and government institutions will be empowered to integrate cutting-edge technologies using augmented intelligence that truly speaks Ukrainian and understands Ukraine. We have a responsibility to bring the benefits of augmented intelligence to the countries we serve — through large language models trained not only on words, but on local context.”

Once operational, the Ukrainian LLM is expected to support applications across public and private sectors. Use cases include legal and regulatory tools as well as solutions for education, finance, and healthcare. Locally trained data is expected to improve relevance and accuracy.

Next Steps and Technical Roadmap

Google Cloud's Krzysztof Kaziów noted the model’s multilingual support and suitability for regional use. Kyivstar will next enhance Gemma for Ukrainian, refine the tokenizer, and train the model on curated datasets. Benchmarks will guide future improvements.

The initiative aligns with VEON’s broader efforts in AI development across its markets, following projects in Kazakhstan and Pakistan.