Poland’s AI future: From gigafactory plans to private sector engagement

XYZ: How is progress on Poland’s AI gigafactory coming along?

Dariusz Standerski, Deputy Minister of Digitalization: In mid-August, we held a meeting with the European Commission to discuss our application, which had been provisionally approved in June. We were encouraged to engage with smaller consortia interested in building their own gigafactories – with a view to potential mergers or collaborations. That is why this week I am meeting with the Czech Minister for Artificial Intelligence to discuss expanding the Baltic AI Gigafactory and the possibility of a joint project. We were also asked to “take care” of so-called trial applications.

XYZ: What does that mean?

Standerski: At first, I was surprised that the European Commission did not publish the full list of the 76 submitted applications. It seemed unusual. After discussions with the Commission, however, I understood their approach: some of the applications were purely trial exercises or lacked a genuine objective to build an AI gigafactory. In Poland, there were also two such cases – applications submitted by a company and financial investors. There was also an example of a foreign firm that was just entering the Polish market and, perhaps hoping for media attention, submitted an application.

The Commission asked us to engage in dialogue with other applicants, but at the same time emphasized that it would not publish a list of all companies and expects Poland not to do so either. The reason is simple: the applications were submitted for different motivations, and not all of them were serious.

The Fate of Poland’s AI Gigafactory

XYZ: How many of these applications represent real competition for Poland? Initially, the European Commission stated it would fund only six projects.

Dariusz Standerski: Out of 76 applications, we estimate that around 30 were submitted with the genuine intention of building factories. However, considering that some countries submitted multiple applications for the same location, the number of real projects likely to reach the finish line is about 15.

XYZ: So Poland’s chances aren’t one in ten, but closer to one in two?

Standerski: Exactly. That’s why the Commission encourages consortia to naturally merge and form larger projects. Ultimately, about six gigafactories are expected to be built across Europe. I suspect the Commission is also factoring in geography – issues like infrastructure, fiber-optic connectivity, and resource availability are crucial. They want clusters to emerge in different parts of Europe.

XYZ: The Commission seems reluctant to share information about the gigafactory project…

Standerski: That may be surprising, but the logic is simple: the Commission had its own concept, waited for initiatives from member states, and now – after collecting the applications – is finalizing the detailed conditions. In the meantime, we are finalizing our consortium. The goal is to submit the final application at the beginning of next year.

The Ministry of Digitalization on what Poland’s AI Gigafactory will look like

XYZ: What is your vision for the application you will submit together with your partners?

Dariusz Standerski: First and foremost, I have two main objectives for the gigafactory -beyond the obvious one, which is building it.

The first goal, which I have already discussed with the European Commission, is for the submission of the application itself to be just another stage in the consortium’s collaboration. I want the companies to start working together now. On August 27, we have the first steering committee meeting of the planned consortium in Warsaw, and I intend to begin work on joint initiatives then. This includes, for example, mutual use of models – Polish models in other countries and vice versa – as well as cooperation on implementations, including in public administration. We have plenty of ideas, and the companies will certainly bring their own, as for many of them this will be their first engagement with the market of another country.

I would like several initiatives to be up and running by December, through which companies and public institutions begin genuine collaboration. This way, the application for the gigafactory will not be the first time everyone meets, but a subsequent stage of already ongoing cooperation.

The second goal is to create an initiative and a joint investment even if we do not receive funding from the European Commission. Of course, on a smaller scale, but I want the gigafactory to be built even without EU funds.

XYZ: Why?

Standerski: Because under the new public-private partnership regulations – which we are awaiting – such a formula could simply be economically viable. The public and private contributions would create shared infrastructure, managed like a company, proportionally to the capital invested.

That is why I believe our greatest failure would not be failing to secure EU funding, but rather if, after months of work, we parted ways and said: “See you at the next application.”

XYZ: What are the concrete expectations for companies joining the consortium?

Standerski: Once we receive more detailed information on the conditions of the Commission-funded partnership, we will send inquiries to companies regarding financial contributions. Based on that, we will determine later participation in procurement contracts or ownership of computing power. The level of private contributions will also determine the amount of EU co-financing we seek – if the 17 percent threshold, previously mentioned by the Commission, is maintained.

Private companies and a state-led investment

XYZ: Business representatives had many concerns about the process, which we recently reported on. The main issue was precisely the lack of concrete information regarding their potential financial involvement.

Dariusz Standerski: I want to clarify all of this. The conversation about large – significantly larger than before – investments in artificial intelligence began on February 11 in Paris. That was when the President of the European Commission announced the launch of the Invest AI program, of which AI gigafactories became a part. In practice, the entire project we are discussing officially started on February 11.

On April 9, the Commission published the framework of the Invest AI program and, at the same time, announced the Call for Expression of Interest for AI gigafactories. That same day, we at the Ministry of Digitalization began work on preparing Poland’s application. From April 9 to June 12, we conducted consultations – we announced an internal call on the ministry’s website and organized a meeting dedicated to how the Polish application should be structured, especially regarding financing and project structure.

XYZ: Why was that necessary?

Standerski: Because the Commission allowed applications to be submitted but had not yet specified the exact rules regarding funding or management. As a result, we proposed several different options.

XYZ: Were private companies involved in that meeting?

Standerski: No, that was an earlier stage. To be precise: before submitting the application, we organized a large meeting at the ministry in May. We discussed how our application should look – what the financing structure, ownership, and project management would be.

The last open consultations were held on June 12. As a result, between June 12 and 20, we finalized our submission. On June 20, it was sent – it was a preliminary document, prepared after a series of meetings with interested organizations and entities.

It is worth emphasizing that at that time, we had no regulations from the Commission. This was probably the first time the EC sent an invitation to member states and essentially said: “It’s up to you how you envision this. Tell us how to make it happen.”

The European Commission and its unusual approach

XYZ: Typically, EU projects are highly formalized…

Dariusz Standerski: This was a complete reversal of the European Commission’s usual approach. The Commission made it very clear: you have the conditions – maximum 35 percent public funding and at least 65 percent private. Show us how you want to organize it. And that’s exactly what we had to calmly explain to some companies – we did so during meetings and debates organized in the media.

At one such event, company representatives told me they wanted to get involved in the gigafactory project but didn’t know the specific conditions. I explained that it was up to us to define those conditions, because that is precisely what the Commission was asking. Their response: “Ah, we didn’t realize. In that case, we’ll send our proposal.” And indeed, they did.

On June 30, the Commission announced that it had received 76 applications. On July 9, during a meeting of the AI subcommittee, I presented the details of our application. A week later, on July 16, the Commission unveiled preliminary regulatory changes regarding gigafactories and explained how the project would be implemented in subsequent steps.

This was likely based on the submitted applications. Some of Poland’s proposals were adopted. For example, regarding funding, the Commission declared it would cover a maximum of 17 percent of project costs, while allowing member states greater flexibility in their public funding contributions. This aligned with our proposal – we had argued that a rigid 35 percent ceiling made no sense. What if we wanted to invest 37 percent? The Commission backed down. Our proposal that access rights (so-called compute time) be allocated proportionally to financial contributions was also accepted. This means the private sector already sees tangible benefits from its involvement at an early stage.

Additionally, we are still working on how state contributions should be treated. Will this be limited to budgetary funds, or will it also include, for example, infrastructure guarantees or co-financing via BGK? How will the Commission determine whether this counts as public funding or not? This still needs clarification. It is also important that the Commission agreed to include prior investments in the gigafactory project. This is crucial for us, as we want to acquire computing power even before the factory is built.

XYZ: Are we talking about investments in Kraków and Poznań, the first AI factories?

Standerski: No, these are separate projects. The purchases are intended for the future gigafactory. We want the equipment bought in the near term to be temporarily used in existing centers, and once the gigafactory is operational, to consolidate everything in one location, achieving economies of scale. The Commission has provisionally accepted this, and we are waiting for concrete regulations.

Polish proposals for changes to the EU project

XYZ: I understand – so this isn’t about past investments, but about purchases for the gigafactory before the applications have even been assessed?

Dariusz Standerski: Exactly. We want computing power to be temporarily stored and used in existing centers, and once the gigafactory is operational, to consolidate everything in one place to achieve economies of scale. That’s why it’s important for us that this be recognized. The Commission has provisionally accepted it, which I’m very pleased about. We are now waiting for concrete regulations, which are expected soon.

From July 28 to August 11, we held a call to expand the consortium to include more entities – both commercial and academic or organizational. Interest was enormous: we received 275 applications, including 148 from Poland, 72 from Lithuania, 24 from Latvia, 15 from Estonia, and single submissions from Luxembourg, the United Kingdom, Ireland, Spain, Belgium, Sweden, Finland, France, Malta, Germany, and Norway.

As a result, the consortium today is much broader than we initially anticipated, when it included entities from only four countries. That is why we proposed establishing a steering committee to organize the applications and select participants according to specific categories. We divided them into seven “baskets”: four investment-related – technology providers, SaaS investors, internal investors needing infrastructure for their own use, and financial investors such as funds or capital market participants – and three non-investment-related: HPC centers, the public sector, and the AI ecosystem.

To be clear: we do not expect financial contributions from the three non-investment baskets – and we will not request them in the next stage of the application process. This is primarily about organizing the process. From each basket, six representatives will participate – three from Poland and one each from Lithuania, Latvia, and Estonia. This will create a group of around forty people who will coordinate work on the application itself.

The future of the AI consortium

XYZ: Will all the companies that applied join the consortium?

Dariusz Standerski: No, we are conducting a verification process. We are checking debt records, credit registries, and any potential illegal activities. At this stage, that’s all we are doing.

XYZ: What if an entity connected, for example, to Russian capital – or otherwise “undesirable” – applies?

Standerski: That falls under the responsibility of the relevant specialized authorities. They are the ones we have asked to provide support.

XYZ: Among the current applicants, are there participants who were part of the application submitted in June?

Standerski: For the most part, yes. The current call had two objectives: to identify new partners and, at the same time, to confirm the engagement of those who participated earlier. Some entities that were with us in June decided not to apply – and I fully understand that. However, significantly more have joined. After the verification I mentioned, we will have a complete list and will announce it officially.

The investor group is closed

XYZ: So the 275 applications represent the closed group of potential investors?

Dariusz Standerski: Yes, for now. Any addition of new partners will be subject to the decision of the Steering Committee.

XYZ: The EUR 3 billion valuation of the gigafactory you mentioned in discussions with companies – are those just preliminary estimates?

Standerski: Exactly. The final amount will depend on private contributions. If there is a shortfall, we can fill it with public funds and only then apply for the European Commission’s contribution. But it also works the other way – if we merge with other consortia or other countries join us, we will need to scale up the project to reflect their financial involvement.

The Polish-Czech factory project: When will details emerge?

XYZ: When will it be clear whether you will actually join forces with, for example, the Czechs?

Dariusz Standerski: I have a meeting with the Czech minister this week. After that, we plan a series of discussions with the European Commission. I think the matter will be clarified by October.

XYZ: If a merger with the Czechs does happen, how significantly would it change the project?

Standerski: That’s one of the questions I will ask my Czech counterpart. We have previously discussed their application and mutually noted in our documents that we want to cooperate. They recorded collaboration with us in their application, and we did the same with them. Now it remains to specify the scale of their project.

XYZ: And what happens next?

Standerski: We will evaluate their finances, which companies are involved, and simply compare the two models to assess the potential.

XYZ: So you are also holding parallel meetings with the European Commission to monitor the process? Only then will you approach companies regarding their potential financial involvement?

Standerski: Exactly.

The concept of Poland’s AI Gigafactory

XYZ: How will the computing power in the gigafactory be allocated? If one company invests, say, PLN 1 million and another PLN 100,000, will the first get ten times more computing power?

Dariusz Standerski: Simplifying: yes. It’s a mechanism familiar from commercial companies – proportionality of contribution and rights. But above all, we must remember the mission of AI gigafactories. They differ from ordinary AI factories not just in scale. Centers like the AI factory in Poznań will adapt models to the needs of specific companies. The gigafactory, however, is meant to be Europe’s answer to the American Stargate project. It will be a hub for training advanced, multimodal models, important for entire industries and systems.

For example, today, to create the next version of PLLuM, the Polish LLM model, we distribute the use of available computing power over several months. In the consortium’s gigafactory – later perhaps a company – a decision could be made that, say, 20 percent of capacity is allocated to training a model simulating biochemical processes or analyzing satellite images. That project would then be executed. It will operate like a company jointly managing resources, not like a university or parliament. The scale of resources, however, will be unprecedented in Poland.

Keep in mind that Poland trained PLLuM for a year. We also have Bielik, developed by a highly engaged community. Estonia is developing its Tilde model. That’s about it for larger models in our region. With the gigafactory, we will be able to train dozens of models on much larger datasets – not only linguistic but also multimodal.

The trained model will then be made available to consortium companies, which can fine-tune it for their own needs, implement it in their AI factories, or offer related services. Alternatively, they can train their own models. This will be a real commercialization mechanism.

XYZ: So, for example, if Allegro (Poland's dominant e-commerce platform) or Orange were in the consortium, they could use a new version of PLLuM? And if the project generates revenue, profits would go to the company in which these firms also hold a stake?

Standerski: Exactly.

XYZ: When will you start gathering the infrastructure necessary for the gigafactory?

Standerski: We are waiting for guidance from the European Commission on whether funds invested in the gigafactory before the application is approved can be counted. If the answer is positive, we will begin the procurement process next year.

XYZ: The gigafactory only in two locations?

Standerski: There have been reports that the gigafactory would be dispersed—five locations in Poland. Is that current?

The European Commission has indicated a maximum of two locations. We are now evaluating this in terms of technical and security considerations. One location may be more efficient, but is it safe from a physical and energy perspective? That is what we will weigh together with the consortium.

XYZ: And if the Czechs join, does the situation become more complicated? They would probably want something in their own country as well…

Standerski: It’s hard to say whether it’s a complication or rather a new opportunity.

Energy for the Gigafactory

XYZ: What about securing access to energy? The gigafactory will be a massive consumer…

Dariusz Standerski: That’s true – it will be a data center several times larger than any currently operating in Poland. That is why we are working with energy companies to ensure a stable supply. We are relying on simulations of new resources and planning connections to the largest facilities. We are clear: the factory must be located where there is strong energy and network infrastructure. It cannot be “in the middle of a field” – it has to be a place with established energy and telecommunications support. And a good energy network exists where leading research centers are already operating.

XYZ: For example, in the Bełchatów region? (Bełchatów is a medium-sized town in central Poland that’s synonymous with Europe’s largest lignite (brown coal) mine and coal-fired power plant complex. This industrial giant defines the entire region's identity and economy)

Standerski: Yes, but the gigafactory is a long-term project. We cannot rely on a location whose future is uncertain. Ideally, for example, it could be in Pomerania, where a nuclear power plant is planned. The problem is that it is a multi-year investment, whereas the gigafactory will be implemented more quickly. That is why we must base it on the infrastructure we currently have.

XYZ: So there is no concept of building it in Lithuania or Czechia, where there are nuclear power plants?

Standerski: It’s not that simple. The key issue is energy transmission, not simply proximity to the source. We will analyze every scenario, but the basic plan assumes the gigafactory will be built in Poland. We will see whether it will be in one or two locations. Potentially, a second location could be in another country – and that possibility is also being considered.

Progress on AI factories: Poznań operational in a year

XYZ: What’s happening with the two AI factories you have already announced?

Dariusz Standerski: The investment in Kraków is still under review by the European Commission – the procedure has been extended, so we are waiting. National funding is secured, and we hope for a positive decision from the Commission. In Poznań, the project is already underway. PLN 400 million of public funds has been released, the site prepared, and contracts are in the process of being signed. We are awaiting full delivery of computing power.

XYZ: When will the Poznań center be fully operational?

Standerski: In 12–13 months, by the end of 2026.

XYZ: What will this investment actually change in AI development?

Standerski: It will focus on three main areas: healthcare and diagnostics, automation of production processes, and cybersecurity. We want access to its resources to be as simple as possible – no dozens of forms, with requests submitted remotely. The factory and the entire process of accessing computing power in Poland should operate as efficiently as possible.

XYZ: So, for example, startups or technology companies will be able to use it?

Standerski: Yes. We are also discussing with the European Commission the possibility of making some capacity available as grants – under de minimis aid. That depends on the Commission, because this would effectively be public aid.

XYZ: And Kraków? When might the Commission make a decision?

Standerski: Knowing EU procedures, I can say one thing: deadlines are for member states, not the Commission. If a country is late, it pays penalties. If the Commission is late… well, it’s just late.

XYZ: Right – the one distributing the money decides.

Standerski: Exactly. We’ve already seen serious delays in adopting standards for the AI Act, in the European digital identity wallet project… And then the Commission is surprised that member states haven’t implemented the new regulations. How could they, when the standards were published on July 31 and expected to be implemented from August 2?

Delays on the Commission’s side

XYZ: The question is whether similar delays might occur with the gigafactories…

Dariusz Standerski: We have to be prepared for every scenario. That’s why from the start we set two objectives: we work within the framework of the European Commission, but in parallel, we are preparing for a scenario in which we finance the project independently. In this case, however, the Commission has strong motivation – the current budgetary period is coming to an end. The Commission wants to allocate funding between the current and the next period. And as you rightly noted – the one who distributes the money also has an interest in ensuring it is fully allocated by the end of 2027.

XYZ: Are additional AI factories planned?

Standerski: The project under that name has been completed – we are now talking about gigafactories. Of course, we will support the development of computing capacity in individual centers, but that falls under national funding. The key to AI development in this area is now in the hands of the Minister of Finance and Economy, Andrzej Domański, because we – as the Ministry of Digitalization – have done what we could.

The Finance Minister: Key to AI development

XYZ: How can all of this be financed, given the difficult budgetary situation?

Dariusz Standerski: I am satisfied, because we have managed to secure truly significant funds. We have PLN 400 million for Poznań and nearly PLN 150 million for Kraków. We launched a loan program for companies – covering cloud and AI solutions – worth PLN 2.7 billion, with an interest rate of 0.5 percent. We have secured additional funds and are awaiting the outcome of EU programs worth over PLN 200 million in AI implementation. This should be resolved in the coming weeks, as these are still funds from the current EU budgetary perspective.

In total, we have secured more than PLN 3 billion for AI over the coming years: 2025, 2026, and partially 2027. It is worth noting that these resources focus primarily on computing power and AI development – the foundations of digitalization. Implementation in industry and the wider economy, however, falls under the Minister of Finance and Economy, who controls the budget. Our capacity in this area is limited.

XYZ: You speak of secured funding, but what happens next? It’s already clear that the 2026 budget will be – mildly put – challenging, especially with the president signaling a veto on tax increases. Aren’t you concerned about the future of digital projects in Poland?

Dariusz Standerski: Every additional zloty for AI development depends on decisions made by Minister Domański. The minister recently announced, among other things, the DeepTech fund, so we are waiting for his next moves. Of course, as an economist, I always have budgetary concerns. But to be honest – the thing that really keeps me awake at night is not public funding, but private investment.

We have done everything we promised. There is PLLuM, the AI factory in Poznań exists, Kraków will follow, and we are making our language model available to the administration. The problem is that Poland today has a very low level of private investment. Overall, we are at the bottom of the European Union, and even so, our investment indicators are driven primarily by the public sector. Regarding private investment, Poland ranks second to last in the entire EU. The public sector provides tools and access to infrastructure, but the private sector is not yet contributing its own resources.

And that, in my view, is the greatest challenge. The extent to which the economy succeeds in the era of artificial intelligence will depend on it.

How to Invest in AI?

XYZ: Where do you see the problem? This low level of private investment in GDP has persisted for a decade, with no significant improvement.

Dariusz Standerski: Perhaps it’s a question of risk aversion among Polish entrepreneurs. It may also be a consequence of the prevailing business model. You can see this in cloud solutions. Just a few years ago, Poland was at the back of the EU – cloud adoption was a few, then a dozen percent. And suddenly, in the span of a single year, it jumped from the mid-twenties to around 50 percent.

This showed that once the technology became “everyday” – easily accessible, available via an inexpensive subscription—companies simply started buying it. But in my view, we were late to this race. Companies that implemented cloud solutions early invested in solutions tailored to themselves, unique tools that allowed them to stand out and increase productivity. Those companies gained a competitive advantage.

In Poland, for the most part, we only jumped in once cloud computing became standard. The problem is that most companies simply purchased a standard subscription – say, a few thousand zlotys per year – the same as millions of other customers. They moved to the cloud, but it was not specially tailored to their needs.

The Ministry won’t rely on Big Tech

XYZ: Programmers were saying: “Let’s move to Azure or Google Cloud,” and companies were buying whatever was available on the website. Click, paid, and done?

Dariusz Standerski: It’s hard to say whether such investments actually increased productivity. That’s my concern – when it comes to AI, some companies may wait another two or three years for “click-and-go” solutions. Then they’ll implement them simply into their systems, but the approach will likely be superficial rather than strategic.

XYZ: Perhaps the Ministry should take additional steps to encourage the private sector to invest?

Standerski: All our projects – like the AI factory – will be supported by an information campaign. We want to encourage companies to use the tools available to them, because they are being developed precisely for their needs. The same applies to PLLuM – we are making it freely available; you can simply download it. Some companies are already using it. We are also signing agreements with local governments. Once work on AI policy is complete, we will prepare operational programs for regional centers. We want our models to be adapted and customized for regional needs. The goal is to reach local businesses.

The problem, however, is that even if we reach companies, demonstrate the tools, and provide access, these remain sector-specific solutions. The question is whether this is enough, given that companies need more comprehensive support – not only in AI or cloud computing but also in areas such as taxation, transport, and logistics. All these instruments today are in the hands of the Minister of Finance and Economy.

Is it worth handing the reins to Big Tech?

XYZ: Aren’t you tempted to just strike a deal with the giants – Google, Microsoft – and say: “Here’s the infrastructure, go ahead and use it”?

Dariusz Standerski: Each of these companies has a clear business model and every right to pursue it. Naturally, the solutions they offer are designed to scale to as many clients as possible, rather than being tailored individually.

Theoretically, yes. I can imagine that if, for example, Microsoft could integrate its solutions with mObywatel – Office 365, Copilot, or other tools – it would gladly do so. That would provide access to a huge user base. The same applies to integration with ZUS or tax offices.

I think technology companies would be happy to offer such services. But we’ve already seen this with cloud computing. At first, adoption was slow for several years, and then -in 2023 or early 2024 – suddenly hundreds of thousands of companies signed up for simple subscriptions. That was beneficial for the providers, but was it really beneficial for the companies themselves? Not necessarily. An effective solution should be tailored, addressing a specific business model. If a company is faced with the choice of buying a subscription or not using the service at all, that’s not customization – it’s mass deployment. The worst-case scenario would be forcing businesses to install “one chatbot” from a single company on their website.

XYZ: So your goal isn’t to push ready-made solutions from big tech quickly, but rather to build infrastructure on which tools tailored to Polish companies’ needs can be developed?

Standerski: Exactly. That’s why we are building infrastructure in Poland, and why the AI factory in Poznań exists – so that, for example, a transport company, a toy manufacturer, or an accounting firm can commission solutions suited to their needs. A subscription that’s the same for everyone will never be as effective.

What’s next for the AI Fund?

XYZ: What’s happening with the AI Fund? It made headlines around late 2024 and early 2025, but for some time now there’s been little news.

Dariusz Standerski: The AI Fund is active. The entities forming the Fund’s Council meet to coordinate their strategies and investment programs. This allows us to align our projects. At the Ministry of Digital Affairs, this is very tangible – for example, the programs we are finalizing under European digital development funds were consulted within the AI Fund Council. This ensures we don’t duplicate initiatives from the education sector, the Ministry of Defense, or other institutions. That was the Council’s main goal: to make sure public money is spent once, rather than multiple times in parallel. By the end of the year, we plan to summarize how much funding has been agreed upon and how it has been allocated.

XYZ: So, I understand there won’t be a single AI fund functioning like PFR Ventures, investing directly in startups?

Standerski: Exactly. From the start, the AI Fund was intended to play a coordinating role. We have multiple sources of financing, and we want to reconcile them. We’ve seen cases where two ministries spent money twice on the same project, and the results now sit on a shelf. So this work may not be spectacular, but it is effective. We’ve stopped using public funds inefficiently – which, frankly, are never as abundant as we’d like, and are critical in new technologies.

XYZ: Will this strategy change?

Standerski: No. However, Minister Domański has announced additional resources in this area through the DeepTech fund.

Key Takeaways

1. Poland is actively advancing the Baltic AI Gigafactory project, seeking not only EU funding but also preparing to execute the investment independently if needed. The consortium, led by the Ministry of Digital Affairs, is steadily expanding to include new partners and institutions from Poland and the Baltic region. Discussions are ongoing with potential allies, including the Czech Republic, as well as with the European Commission, which plans to finance only six gigafactories across Europe. Poland’s proposal is among roughly 15 projects realistically under consideration. The final application is scheduled to be submitted at the beginning of 2026.
2. The initiative aims not merely to build infrastructure but to establish enduring public-private collaboration in AI. The consortium will facilitate shared use of AI models, cooperation on deployments across different countries, and development of solutions tailored to industrial and regional needs. Even without EU funding, the project is set to continue under a public-private partnership model. An important element is the early acquisition of hardware and computing capacity, which will be counted toward the gigafactory investment.
3. Despite secured public funding, the main challenge remains the low level of private investment in Poland. Projects such as the AI factories in Poznań and Kraków already have financing and are at an advanced stage, but their long-term success will depend on business engagement. The Ministry is conducting educational and promotional efforts, emphasizing that key investment decisions rest with private companies. A lack of strategic, long-term private investments could ultimately determine Poland’s competitiveness in the era of artificial intelligence.