This article is a part of Poland Unpacked. Weekly intelligence for decision-makers
Next Frontier AI is a pan-European initiative launched in November 2025. Its goal is to connect the public sector, technology institutions, and private capital to build innovations at a global scale—particularly in advanced artificial intelligence.
The initiative itself is funded by SPRIND, a German government agency dedicated to breakthrough innovation. It supports deep-tech startups.
“The ambition of this initiative is very clearly to build something on the scale of OpenAI, Anthropic, or Google DeepMind—but in Europe and for Europe,” says Dr. Johannes Otterbach, an investment advisor at SPRIND and one of the initiators of Next Frontier AI.
Now, Polish companies have an opportunity to join.
Next Frontier AI comes to Poland
The first project linking the initiative with Poland is the Next Frontier AI – WARSAW event, scheduled for April 22. Organizers are bringing together ambitious companies and teams with the potential to deliver high-impact innovation. The best participants may qualify for a program selecting 10 top teams from across Europe.
The stakes are high: up to EUR 3 million (approx. PLN 12.9 million) in non-dilutive funding—meaning no equity is required in return—and the opportunity to raise additional large financing rounds.
“Poland, and Warsaw in particular, has always been an obvious point on the map for me. During my time at OpenAI, I saw firsthand how strong Poland’s contribution to frontier AI is. Many members of OpenAI’s technical teams came from Poland or had Polish roots, including co-founder Wojciech Zaremba. That is no coincidence. Warsaw and other Polish cities have strong traditions in computer science and educate world-class researchers and engineers,” explains Johannes Otterbach.
Poland’s opportunity in AI
The direct initiators of the Polish edition of Frontier AI are Łukasz Alwast (representing the startup and VC ecosystem) and Aleksandra Pędraszewska (formerly with ElevenLabs, now managing the venture fund Vastpoint).
“Next Frontier AI brings European talent closer to capital and infrastructure designed to increase the chances of building teams capable of competing globally and creating alternative solutions. Late? Possibly. But this race still depends on asymmetries in access to opportunities and information, including geographic ones. Our goal is to ensure that Polish talent has real access to such opportunities—not learns about them after the fact,” says Łukasz Alwast.
Aleksandra Pędraszewska highlights a tangible benefit of participating in the program.
“The direct impact is concrete: €125 million (approx. PLN 537.5 million) in funding becomes accessible to Polish teams. Without our participation in the roadshow, many would likely not have heard about this initiative. In a world defined by asymmetries in access to opportunities and resources, this is critical. I am glad that together with Łukasz—both of us self-described ‘ex-expats’—we are bringing back to Poland not only experience but also a network and access to initiatives like this,” she notes.
How Next Frontier AI works
Next Frontier AI fits into a broader wave of European initiatives aimed at strengthening innovation ecosystems, such as the UK’s Sovereign AI project or France’s Bpifrance DeepTech program. SPRIND, however, seeks to leverage not only German ideas and has significantly larger resources than comparable initiatives.
“We are looking both for the next ‘OpenAI’ and for highly specialized technology projects. Today, we do not know where exactly the next frontier of AI will emerge. The current paradigm of scaling large language models has defined an era, but the question we ask is: what comes next?” says Johannes Otterbach.
In practice, the Frontier AI program aims to select ten teams that each propose their own hypothesis on how to achieve the next breakthrough in AI.
“One team may focus on new model architectures that are radically more efficient. Another may concentrate on agent-based systems where thousands of smaller models collaborate. Another may explore embodied AI or world models for industry. Yet another may work on breakthroughs in evaluation methods or data pipelines. This diversity is intentional,” he adds.
Teams will be selected through a series of “roadshow” events, including one in Warsaw. Previous editions have taken place in Munich and Paris, with upcoming events planned in Zurich and Amsterdam. A jury will select the 10 projects, each eligible for €3 million (approx. PLN 12.9 million) in funding.
“Over the next two years, based on a milestone-driven system, the number of teams will be narrowed down to three. These will then receive structured support to raise up to €1 billion (approx. PLN 4.3 billion) each. The goal is for them to evolve into full-fledged frontier AI laboratories. In other words, we are looking for the next frontier AI lab—but the path involves many specialized teams pursuing different visions of what that frontier might be,” Otterbach explains.
How to join Next Frontier AI
The Warsaw event will take place on April 22. The official application process will begin in May. Organizers expect to select the 10 teams and allocate funding in July. Each selected team will receive EUR 3 million (approx. PLN 12.9 million) in non-repayable funding. The top three teams may receive an additional EUR 23.5 million (approx. PLN 101.05 million) each over the following two years.
“The jury will include elite practitioners and founders, such as Pim de Witte from General Intuition. If even two or three teams from Warsaw secure this funding or establish relationships that strengthen the ecosystem, that would represent a meaningful long-term success,” says Łukasz Alwast.
According to Aleksandra Pędraszewska, simply including Warsaw among the host cities is already a significant step.
“If in two years one of the three winning frontier AI labs funded under NFAI has Polish founders, that will be another milestone. ElevenLabs and OpenAI founders have already reshaped the narrative around Poland’s tech ecosystem—from a software back office to leaders in AI research. The next globally successful company with Polish DNA will achieve more than a decade of promotion,” she notes.
A SPRIND representative emphasizes that the concentration of tech talent in the region is substantial.
“The density of talent in cities such as Warsaw, Kraków, Prague, or Bucharest is real and often underestimated by those focused solely on London, Paris, or Zurich. That is why it was crucial for Next Frontier AI to be present in Poland and actively reach teams in this region,” says Johannes Otterbach.
What is happening in Polish innovation
Poland’s deep-tech ecosystem is currently financed through a hybrid model with a clear dominance of public capital. Government and EU funds play a central role—primarily via PFR Ventures and programs such as FENG and earlier Bridge Alfa. These funds supply venture capital firms that then invest in startups. A significant share of transactions involves funds from the PFR portfolio, with deal activity rising year over year. Specialized initiatives are also emerging, such as the PFR Deep Tech program (around PLN 600 million / approx. EUR 139.5 million), which channels capital into sectors including AI, robotics, and cybersecurity.
The second pillar consists of private investors, including local business angels and international VC funds, which account for roughly 40% of market capital. In practice, most larger funding rounds today are mixed (public-private).
“One event or program will not fix the system. In parallel, within the Future Council we are working on structural reforms—such as tax incentives for angel investors, digitization of registries and share trading, and better technology transfer from universities. These are two tracks that must run simultaneously. NFAI without reform is a boost for a few teams; reform without instruments like NFAI builds a system that remains empty for too long. We need both,” says Aleksandra Pędraszewska.
Expert's perspective
How to stimulate Polish innovation
The main barrier remains risk aversion and insufficient access to capital capable of financing deep tech at scale. There is also a lack of effective bridges connecting academia with globally competitive business.
A key priority is providing researchers with access to computing power, which is essential for training frontier AI models. Equally important is the creation of ambitious, long-term R&D projects locally—projects that can offer challenges comparable to those posed by Silicon Valley giants.
The AI race
The search for AI technologies is gaining increasing importance. The term “frontier AI” has become widely used to describe advanced systems operating at the limits of current technological capability.
“What we consider frontier AI is a moving target that shifts quarter by quarter. Those working at that frontier see best where and how fast the field is evolving—especially in leading first-generation AI labs such as Google DeepMind, OpenAI, Anthropic, and NVIDIA, as well as next-generation players like Runway, ElevenLabs, Mistral, or Black Forest Labs,” lists Łukasz Alwast.
In Poland, alongside specialists working for global tech giants, there is also a community of academic researchers and independent programmers.
“We are talking about a narrow group of a few hundred individuals of Polish origin—and, more broadly, tens of thousands worldwide. These individuals have the highest probability of forming teams with a ‘right to win’—capable of competing both technologically and commercially. Success is not just a matter of capital, mindset, or talent, but all three combined, at a scale that founders in the Polish ecosystem have not yet fully faced. Competence is built only through attempts at the right level of difficulty,” he adds.
According to Aleksandra Pędraszewska, building a true frontier AI startup requires not only an idea and technical expertise but also business capabilities.
“Building a frontier AI company today requires both strong research and the ability to close contracts, manage runway, and recruit top engineers. These skills are rarely found in the same person. SPRIND explicitly requires PhD-level research combined with operational experience in its NFAI criteria—and that is no coincidence,” says the Vastpoint partner.
Key Takeaways
- Structural challenges remain despite increasing funding. Although public funding is available and private investment is growing, barriers persist—particularly risk aversion and limited scale-up capital. Experts also point to weak mechanisms linking academia with business and limited access to computing infrastructure. Initiatives like Frontier AI may act as catalysts, but without parallel systemic reforms, their impact will remain constrained.
- Europe’s growing ambition in AI. The Frontier AI initiative, funded by SPRIND, aims to build an ecosystem capable of competing with leaders such as OpenAI and Google DeepMind. It promotes collaboration between public, private, and tech sectors and supports diverse approaches—from new model architectures to agent systems and industrial applications—signaling Europe’s intent to close the technological gap with the United States.
- Poland joins the European AI ecosystem as a key talent hub. Hosting the Next Frontier AI event in Warsaw provides Polish teams with access to significant funding and an international network. Experts highlight Poland’s strong engineering and research base, while noting the need for better access to information and capital. Participation could increase the global visibility of Polish firms and reposition the country from a technology back office to an innovation creator.