The European Commission has put one of its biggest AI sovereignty bets in the hands of Domyn, a Milan company still better known to insiders by its old name, iGenius.
On June 19, the European Commission named EUROPA, a consortium led by Milan-based Domyn, as the winner of its Frontier AI Grand Challenge. The prize is not a cheque in the usual sense. It is access to up to 2.5% of total EuroHPC computing capacity for one year across the bloc's AI-optimised supercomputers, with a target that is both blunt and ambitious: build an open-source frontier model of more than 400 billion parameters across all 24 official EU languages in about twelve months.
That is a serious mandate. OpenAI, Anthropic and Google do not publish neat, comparable parameter counts for their current frontier models, so you should be careful with any table that pretends this is a clean horse race. Still, 400 billion parameters is a large public target, and the language requirement makes the job harder than a simple model-size headline suggests. Finnish, Maltese, Irish and Bulgarian are not edge details when the Commission says all 24 official EU languages are in scope. They are the test.
What Brussels is really buying is a hedge. As Il Sole 24 Ore reported in its coverage of the announcement, the EU's stated motivation is reducing dependence on American and Chinese AI providers. That concern has become harder to wave away in 2026. The Guardian recently wrote about a viral Europe 2031 scenario that gained traction among policymakers after fears that access to a foreign AI model could be restricted. You do not have to buy every apocalyptic line in that scenario to see the point. If your courts, ministries, hospitals and public services run on models controlled elsewhere, your sovereignty has a hole in it.
Here's the thing about Domyn: until recently, most people outside European enterprise AI knew it as iGenius, if they knew it at all. Uljan Sharka founded the company in Milan in 2016 after arriving in Italy as a teenager from Albania, learning the language and building a career around enterprise software. The company made its name selling AI systems into regulated sectors such as financial services, government and heavy industry, where data residency and auditability are not branding lines. They are procurement requirements.
That background explains why Domyn fits the Commission's mood. A consumer chatbot company would be the wrong political signal. A Milan group pitching sovereign AI for institutions sounds much closer to what Brussels wants to prove: that Europe can build useful models without handing the stack to San Francisco or Shenzhen.
The Commission said it judged Domyn's proposal on strategic vision, execution capability and potential impact. The EUROPA consortium also includes Fraunhofer-Gesellschaft, the German applied research organisation. That part matters. Fraunhofer is not decorative letterhead. It brings research depth and institutional infrastructure that a company of roughly 150 people cannot credibly claim on its own.
Sharka told Reuters the model will launch within a year. You can read that as confidence, or as the only answer you can give after the European Commission hands your company the continent's flagship AI project. Frankly, it is probably both. Training a 400-billion-parameter multilingual model from scratch in twelve months, on shared supercomputing capacity, through a consortium that has to coordinate across institutions, would stretch a much larger company.
The timeline is the part you should watch. The Frontier AI Grand Challenge launched in February 2026, only four months before Domyn was announced as the winner. That is fast for a project meant to carry this much political weight. Speed can be useful. It can also hide weak assumptions, especially when public institutions are under pressure to show they are not falling behind.
Europe does have pieces worth taking seriously. EuroHPC gives it a route to compute. Fraunhofer and other research institutions give it deep technical benches. The AI Act gives public-sector buyers a language for auditability and accountability that black-box foreign systems may struggle to satisfy. What Europe has lacked is not another strategy document. It has lacked a credible operator to pull the pieces into something people can actually use.
Domyn is now volunteering to be that operator. Its rebrand, its sovereign AI pitch and its positioning against OpenAI and Anthropic all point to this moment. That kind of bet concentrates the mind. It also concentrates the risk. If EUROPA ships a real multilingual model that public institutions can inspect and deploy, Europe gets more than a technical demo. If it slips into another funded initiative with a polished announcement and limited adoption, the continent will still be dependent on the same foreign platforms it says it wants to escape.
For now, the most honest verdict is simple: Brussels has made the right problem impossible to ignore, and it has handed the execution to a company that still has to prove it can carry the weight.
Also read: Waymo registers a German subsidiary and starts recruiting in Berlin and Munich • SpaceX's $25 billion bond frenzy is the clearest sign yet that markets have lost the plot • Unconventional AI raises $475 million at a $4.5 billion seed valuation to cut AI's power bill by a factor of 1,000
