UNIT.City — місце, де люди працюють... КРАЩЕ! Обирай свій простір просто зараз 👉
Наталя ХандусенкоAI Eng
27 January 2026, 13:01
2026-01-27
The start of the national LLM is being slowed down by data collection: what is the problem?
The launch of the national LLM was postponed to the spring of 2026, as the Ministry of Information and Communications faced a data collection problem, which consists of two parts: legal and technical.
The launch of the national LLM was postponed to the spring of 2026, as the Ministry of Information and Communications faced a data collection problem, which consists of two parts: legal and technical.
This was stated by Oleksandr Bornyakov, acting Minister of Digital Transformation, in an interview for DOU and the YouTube channel "UT-2".
The legal part of the problem lies in the fact that the ministry, as a state body, cannot simply automatically collect or remove data that is protected by someone's intellectual property rights, unlike private structures.
“We received a clear ‘hard no’ from our partners regarding the use of certain data sets. If we receive even one lawsuit, the entire project will fall apart,” explains Bornyakov.
Therefore, the ministry cannot take such a risk, since state services are built on this model. To this end, a legal framework for obtaining consents is currently being created.
"We want to adopt a norm: if the information is public and posted on the website in open access, LLM can use it for training," notes the interim head of the Ministry of Digital Affairs.
In addition to the legal part, there was a technical delay. This concerned the creation of the team.
"It's quite difficult to hire IT people now. Kyivstar helped a lot here. As a partner, they took over part of the processes and even involved people from their team. In the end, we managed to form a team," says Bornyakov.
There was also a delay in choosing the platform, but ultimately the basis will be Google's Gemma .
“We will take all the data we have — books, archives — and feed them to the models. Now our own tokenizer is almost complete,” adds the head of the Ministry of Digital Affairs.
Special tests are currently being developed that will demonstrate the quality of the model's performance. After training, they plan to release it.
Ukrainian answer ChatGPT. How Kyivstar and the Ministry of Digital Economy will build a national LLM for Ukraine: insights and international AI experience VEON
Як нейромережі бачать вільну та незалежну Україну? Тест dev.ua
Нейронні мережі для генерації зображень бачать світ по-своєму, їхню логіку зрозуміти часом зовсім неможливо. Але таки хочеться. На честь Дня Незалежності України редакція dev.ua вирішила провести невеликий експеримент.
Ми задали чотирьом різним нейронним мережам п’ять однакових запитів: «прапор України», «День Незалежності України», «український Крим», «перемога України» та «українці». Отриманими результатами ми ділимося з вами нижче.
У TikTok тепер можна генерувати фон за допомогою нейромережі. Ми протестували її та ділимося результатами
У TikTok з’явилася нова функція «Розумний фон». З її допомогою як фон для тіктоків можна підставляти згенеровані нейромережею зображення. Редакція dev.ua протестувала цю технологію і ділиться своїми враженнями.