Scroll Top
  • Home
  • News
  • Suspilne Ukraine provides news archive to train Ukraine’s national large language model
Suspilne Ukraine provides news archive to train Ukraine’s national large language model

Suspilne Ukraine provides news archive to train Ukraine’s national large language model

Поділитись

Suspilne’s news reports, feature stories and interviews will be used to train Siaivo, Ukraine’s national large language model. Suspilne Ukraine has joined the initiative as one of the partners providing licensed access to verified editorial content.

The team sees this step as a natural extension of Suspilne’s mission: high-quality journalism — including news, reports, interviews and special projects covering events around the world and across Ukraine — should become part of the country’s information infrastructure.

Khrystyna Havryliuk, General Producer of News and Current Affairs at Suspilne Ukraine, emphasised:

“Suspilne’s news archive is not merely a dataset for machine learning. It is also a linguistic infrastructure built on original Ukrainian-language content. If we want AI models to understand the Ukrainian context and work with reliable information, they need to be trained on precisely this kind of content.”

Kyrylo Yesin, Head of Suspilne’s Digital Product Development Team, said:

“This initiative provides an opportunity to enhance the value of Suspilne’s digital content: it will become part of a sovereign language model and help lay the foundation for a high-quality digital experience for every Ukrainian.”

Under the partnership agreement, Suspilne is granting the State Enterprise Diia a non-exclusive licence to use its archive of publications. The archive contains more than 764,000 items and includes headlines, article texts and metadata, such as categories, tags, information about authors and publication times.

The parties are also considering expanding the volume of content that will be used to train Ukraine’s national language model in the future.

The Siaivo national LLM initiative is a project led by the Ministry of Digital Transformation of Ukraine, with the participation of the State Enterprise Diia and Kyivstar. It aims to develop a large language model trained on Ukrainian-language data and tailored to the needs of government, businesses and society.

The initiative seeks to reduce Ukraine’s dependence on foreign language models in sensitive areas, including public services, education and media, and to ensure that the Ukrainian language is properly represented in AI systems, where it is currently considered a low-resource language. Ukraine’s national large language model will be trained on data free from propaganda and disinformation.

Suspilne Ukraine is an independent media company with a strong presence across all platforms: the TV channels Pershyi, Suspilne Kultura, Suspilne Sport, and a national network of local channels; the radio stations Ukrainske Radio, Radio Promin, Radio Kultura, Radiotochka. Find only verified news on suspilne.media, as well as on our national and local digital platforms. We broadcast in minority languages, represent Ukraine at the Eurovision Song Contest, develop the children’s platform Brobaks, and train media professionals through the Public Media Academy. Our Suspilne Mediateka platform offers unique video and audio archives dating from the 1920s to today. Defending freedom in Ukraine. 

Поділитись
Suspilne Ukraine provides news archive to train Ukraine’s national large language model

Suspilne Ukraine provides news archive to train Ukraine’s national large language model

Поділитись

Suspilne’s news reports, feature stories and interviews will be used to train Siaivo, Ukraine’s national large language model. Suspilne Ukraine has joined the initiative as one of the partners providing licensed access to verified editorial content.

The team sees this step as a natural extension of Suspilne’s mission: high-quality journalism — including news, reports, interviews and special projects covering events around the world and across Ukraine — should become part of the country’s information infrastructure.

Khrystyna Havryliuk, General Producer of News and Current Affairs at Suspilne Ukraine, emphasised:

“Suspilne’s news archive is not merely a dataset for machine learning. It is also a linguistic infrastructure built on original Ukrainian-language content. If we want AI models to understand the Ukrainian context and work with reliable information, they need to be trained on precisely this kind of content.”

Kyrylo Yesin, Head of Suspilne’s Digital Product Development Team, said:

“This initiative provides an opportunity to enhance the value of Suspilne’s digital content: it will become part of a sovereign language model and help lay the foundation for a high-quality digital experience for every Ukrainian.”

Under the partnership agreement, Suspilne is granting the State Enterprise Diia a non-exclusive licence to use its archive of publications. The archive contains more than 764,000 items and includes headlines, article texts and metadata, such as categories, tags, information about authors and publication times.

The parties are also considering expanding the volume of content that will be used to train Ukraine’s national language model in the future.

The Siaivo national LLM initiative is a project led by the Ministry of Digital Transformation of Ukraine, with the participation of the State Enterprise Diia and Kyivstar. It aims to develop a large language model trained on Ukrainian-language data and tailored to the needs of government, businesses and society.

The initiative seeks to reduce Ukraine’s dependence on foreign language models in sensitive areas, including public services, education and media, and to ensure that the Ukrainian language is properly represented in AI systems, where it is currently considered a low-resource language. Ukraine’s national large language model will be trained on data free from propaganda and disinformation.

Suspilne Ukraine is an independent media company with a strong presence across all platforms: the TV channels Pershyi, Suspilne Kultura, Suspilne Sport, and a national network of local channels; the radio stations Ukrainske Radio, Radio Promin, Radio Kultura, Radiotochka. Find only verified news on suspilne.media, as well as on our national and local digital platforms. We broadcast in minority languages, represent Ukraine at the Eurovision Song Contest, develop the children’s platform Brobaks, and train media professionals through the Public Media Academy. Our Suspilne Mediateka platform offers unique video and audio archives dating from the 1920s to today. Defending freedom in Ukraine. 

Поділитись
Privacy Preferences
When you visit our website, it may store information through your browser from specific services, usually in form of cookies. Here you can change your privacy preferences. Please note that blocking some types of cookies may impact your experience on our website and the services we offer.