Kyivstar, a leading Ukrainian telecom operator and a subsidiary of VEON Ltd., and the Ukrainian Ministry of Digital Transformation have announced a partnership to develop Ukraine’s first national large‑language model (LLM). The project will use Google’s open‑source Gemma 3 model, fine‑tuned on Ukrainian language data and deployed on Google’s Vertex AI platform, with all training and inference performed on servers located within Ukraine to satisfy data‑sovereignty and security requirements.
The choice of Gemma—a lightweight, open‑source generative AI model released by Google DeepMind in early 2025—allows Ukraine to customize the LLM for local dialects, legal terminology, and cultural references while keeping the model’s code and data under national control. Vertex AI provides a managed environment for model training, scaling, and monitoring, enabling Kyivstar to maintain operational oversight and to integrate the LLM into its existing digital services portfolio.
This initiative fits into VEON’s broader “AI1440” strategy, which seeks to launch localized LLMs across its markets to drive customer engagement and new revenue streams. Kyivstar has already committed $1 billion to digital investments in Ukraine between 2023 and 2027, and the partnership is expected to accelerate the rollout of AI‑powered services in public and private sectors, from regulatory analysis to education, finance, and healthcare applications.
By building a sovereign AI capability, Ukraine aims to reduce reliance on foreign proprietary systems and to position itself as a regional leader in AI by 2030. The collaboration also strengthens ties between Ukraine and the United States, as the project leverages U.S. technology while ensuring that sensitive national data remains within Ukrainian borders.
The content on BeyondSPX is for informational purposes only and should not be construed as financial or investment advice. We are not financial advisors. Consult with a qualified professional before making any investment decisions. Any actions you take based on information from this site are solely at your own risk.