Home

Artificial intelligence and Polish science

Where AI is researched and developed in Poland: Polish language models, AI factories and supercomputers, research institutes, government and PFR programmes, and degree programmes with AI in their name.

What AI means for science

In science, artificial intelligence (AI) is three things at once. It is a subject of research – new machine-learning methods, language models, computer vision. It is a tool – it helps analyse data in medicine, astronomy or climate science. And it is infrastructure – training large models requires supercomputers and datasets that a single team usually does not have.

The Polish classification of science has no separate discipline called “artificial intelligence”. AI research is mostly carried out in computer science (natural sciences) and information and communication technology (engineering and technology). In the 2022–2025 evaluation they were assessed at 11 and 48 institutions respectively; 4 of them received the top A+ category.

The Polish AI ecosystem

It is easiest to describe in five layers. The first is people: according to the POL-on register, universities run 67 degree programmes with AI, machine learning or data science in the name, at 31 institutions. The second is models and data: the Polish-language models PLLuM and Bielik. The third is computing power: PLGrid supercomputers and two AI factories, in Poznań and Kraków. The fourth is research institutes, such as the IDEAS Research Institute. The fifth is policy and money: government strategies, AI HUB Poland and programmes of the Polish Development Fund (PFR).

The layers are connected. The same universities and institutes that train specialists build language models on academic supercomputers, and public administration uses the resulting models – PLLuM works as an assistant in the mObywatel app.

How to use this page

Below we describe the most important initiatives, with dates, amounts and partners as given in the sources. Universities and institutes link to their pages in the catalogue, with degree programmes, research categories and the right to award degrees. At the end you will find degree programmes with AI in their name and the institutions with the highest categories in computer science.

Strategy and coordination

The state sets out its goals for AI in strategy documents and builds shared infrastructure for administration and science.

Policy for the Development of Artificial Intelligence in Poland from 2020

The government document in force, adopted by a resolution of the Council of Ministers. It has six pillars, including “AI and science” and “AI and education”.

  • Resolution No. 196 of the Council of Ministers of 28 December 2020 (Monitor Polski 2021, item 23)
  • Pillars: society, innovative companies, science, education, international cooperation, public sector

Source: Uchwała nr 196 Rady Ministrów z 28 grudnia 2020 r. w sprawie ustanowienia „Polityki dla rozwoju sztucznej inteligencji w Polsce od roku 2020” (M.P. 2021 poz. 23)

Policy for the Development of Artificial Intelligence in Poland until 2030

A new document prepared by the Ministry of Digital Affairs. It announces, among other things, a coordinated ecosystem based on PLGrid and AI factories, deployment maps for key sectors and closer cooperation between science and business.

  • Draft resolution of the Council of Ministers (ID260) published on 16 April 2026
  • Among the plans: a Baltic AI GigaFactory

Source: Ministerstwo Cyfryzacji · Prawo.pl

AI HUB Poland

The government portal on artificial intelligence and a project for a central centre supporting the development and deployment of AI in the state. The project “AI HUB for public administration” brings together ministries, research institutes and universities and continues work on Polish language models from the HIVE AI project.

Source: AI HUB Poland · Centrum Projektów Polska Cyfrowa · CRN

Polish language models

A large language model is a program that has learned language from a huge collection of texts and can answer, summarise or write. Polish models are built for the Polish language and Polish uses.

PLLuM

A family of Polish-language models developed on commission from the Ministry of Digital Affairs – first by the PLLuM consortium, since 2025 by the HIVE AI consortium. Training data were used in line with copyright law, part of them licensed directly from their owners. In late 2025 PLLuM was added to the mObywatel app as an assistant that helps find regulations and handle official matters.

Source: ACK Cyfronet AGH · NASK

Bielik

A family of Polish language models created by the SpeakLeash Foundation in cooperation with ACK Cyfronet AGH. The models are open – they can be used privately and commercially; researchers at Polish universities can access them through the PLGrid portal.

  • Sizes: from 1.5 to 11 billion parameters; first widely released model: Bielik-11B-v2
  • Trained on the Helios supercomputer at AGH

Source: ACK Cyfronet AGH

Computing power: supercomputers and AI factories

Training models requires thousands of graphics processors (GPUs). AI factories are computing centres co-funded by the European Union (EuroHPC) and open to science, companies and public administration.

Gaia AI Factory (Kraków)

The second Polish AI factory, led by ACK Cyfronet AGH, in a consortium of 11 partners – including the University of Warsaw (ICM), Gdańsk University of Technology (TASK) and Wrocław University of Science and Technology (WCSS). It is to support, among others, public administration and the space sector, for example the analysis of satellite data.

  • Budget: about EUR 70 million (about PLN 300 million), half from Poland and half from the EU
  • Gaia supercomputer: more than 1,000 GPU accelerators, integrated with PLGrid; the project started in May 2026

Source: Kraków dla biznesu

PLGrid and the Helios supercomputer

The national computing infrastructure for science. The Helios supercomputer at ACK Cyfronet AGH was used to train, among others, the Bielik and PLLuM models.

  • Access for researchers through the PLGrid portal

Source: ACK Cyfronet AGH

Research institutes

Besides universities, AI is developed by research institutes that combine research with applications in the economy and the state.

IDEAS Research Institute

An institute established by a regulation of the Council of Ministers for advanced AI research and its applications in the economy and administration, with a focus on defence, health and public administration. It is to cooperate with the ELLIS network, among others, and support doctoral students and young researchers. It is headed by Prof. Piotr Sankowski.

  • Regulation adopted on 25 February 2025; supervision: Ministry of Digital Affairs and Ministry of National Defence
  • Catalogue page: IDEAS Research Institute

Source: Ministerstwo Cyfryzacji

Polish Development Fund: deployment and investment

PFR links AI with the economy: it helps companies deploy solutions and invests in technology companies, including those that grow out of research.

PFR AI Competence Centre

A PFR initiative supporting the deployment of artificial intelligence in Polish companies; PFR also promotes deployments of the Polish PLLuM model.

Source: PFR · PFR

PFR Deep Tech

A fund-of-funds programme run by PFR Ventures in partnership with the Ministry of National Defence. The money will go to 3–5 venture capital funds, which will select about 50–80 companies in AI, robotics, cybersecurity, space and quantum technologies, including dual-use solutions.

  • PLN 600 million: half from PFR, half from private and institutional investors

Source: ReBiznes

Degree programmes with AI in the name

Programmes whose name includes artificial intelligence, AI, machine learning or data science, according to the POL-on register: 47 names, 67 programmes at 31 institutions. AI is also taught within programmes with other names, such as computer science.

Top categories in computer science

Institutions with category A+ or A in the 2022–2025 evaluation in the disciplines where most AI research is done.

InstitutionCityDisciplineCategory
Akademia Górniczo-Hutnicza im. Stanisława Staszica w KrakowieKrakówinformation and communication technologyA+
Akademia Górniczo-Hutnicza im. Stanisława Staszica w KrakowieKrakówcomputer scienceA+
Uniwersytet Jagielloński w KrakowieKrakówinformation and communication technologyA+
Uniwersytet WarszawskiWarszawacomputer scienceA+
Instytut Badań Systemowych Polskiej Akademii NaukWarszawainformation and communication technologyA
Instytut Chemii Bioorganicznej Polskiej Akademii NaukPoznańinformation and communication technologyA
Instytut Informatyki Teoretycznej i Stosowanej Polskiej Akademii NaukGliwiceinformation and communication technologyA
Instytut Podstawowych Problemów Techniki Polskiej Akademii NaukWarszawainformation and communication technologyA
Instytut Podstaw Informatyki Polskiej Akademii NaukWarszawainformation and communication technologyA
Naukowa i Akademicka Sieć Komputerowa - Państwowy Instytut BadawczyWarszawainformation and communication technologyA
Politechnika CzęstochowskaCzęstochowainformation and communication technologyA
Politechnika GdańskaGdańskinformation and communication technologyA
Politechnika LubelskaLublininformation and communication technologyA
Politechnika PoznańskaPoznańinformation and communication technologyA
Politechnika ŚląskaGliwiceinformation and communication technologyA
Politechnika WarszawskaWarszawainformation and communication technologyA
Społeczna Akademia Nauk z siedzibą w ŁodziŁódźinformation and communication technologyA
Uniwersytet im. Adama Mickiewicza w PoznaniuPoznańinformation and communication technologyA
Uniwersytet Jagielloński w KrakowieKrakówcomputer scienceA
Uniwersytet Mikołaja Kopernika w ToruniuToruńinformation and communication technologyA
Uniwersytet WrocławskiWrocławcomputer scienceA
Uniwersytet WSEI w LublinieLublininformation and communication technologyA

Categories 2022–2025: list published by the Ministry of Science and Higher Education on 22 July 2026. Categories are awarded by decisions of the minister and may change after a request for reconsideration. Categories 2017–2021: open RAD-on data.

Sources

Sources are in Polish unless stated otherwise. Checked: 2026-11-06