Warning: this post is written by academics and past academics — specialists in EU innovation funding schemes and calls. As such, it will, unfortunately, contain long sentences and the passive voice. Some habits cannot be unlearned.
The EU has launched a new wave of EU innovation funding — and this post looks at what it actually offers startups, SMEs and academics. A note on stance before we start: this blog scrutinises Europe’s innovation agenda precisely because we want it to succeed. Applause is free; attention is the compliment.
Last year, we wrote about Mario Draghi’s competitiveness report and its central claim: Europe produces world-class research and then watches other continents commercialise it. We also looked into how this claim resonates for the post-Brexit UK research landscape. Draghi called for the EU to invest seriously in its own technological base — including its own AI development — rather than remaining a customer of everyone else’s.
At the time, it was fair to ask whether any of it would survive contact with the Brussels machine. Twelve months on, the honest answer is: more than sceptics expected.
The AI promise, one year later
In February 2025, Commission President von der Leyen launched InvestAI, a plan to mobilise €200 billion for AI across the EU, with €20 billion ring-fenced to build up to five AI gigafactories: facilities with 100,000+ advanced AI chips each, designed to train frontier-scale models on European soil.
Is it moving fast? By EU standards, remarkably. In January 2026, the Council amended the EuroHPC regulation to mandate the gigafactory programme, and the official call for proposals is expected this summer. An informal call for interest drew 76 proposals across 16 Member States — consortia involving Deutsche Telekom, Scaleway, Nokia and others are already jockeying for position. Has it been delayed? Also, yes — the formal call slipped from late 2025. This is still Brussels. But the direction of travel is unmistakable: the Draghi agenda didn’t die in a drawer. It became a work programme. Several work programmes, in fact.
Which brings us to December’s document
In December 2025, the Commission adopted the Horizon Europe Work Programme 2026–2027 for European Innovation Ecosystems — the funding blueprint that operationalises the EU Startup and Scaleup Strategy. It cites Draghi by name, and it exists to fix the exact failure he diagnosed: European researchers are poorly connected to the networks of universities, startups, corporates and venture capital that turn knowledge into companies.
Most coverage of these programmes focuses on the headline money. I read them differently — through an IP lens — because buried in the plumbing are three developments that matter directly to the two audiences of our report, IP Law: A Global Overview — 2026 Edition: companies with ideas, and academics thinking about commercialising theirs.
1. The EU IP Helpdesk lives on — with a sharper edge
The free European IP Helpdesk, a first-stop resource for EU-funded projects and internationalising SMEs, is being extended to September 2028, with a new procurement round and a €2.6 million budget planned for late 2026.
The interesting part is the expanded remit. Beyond the familiar training and cross-border advice, the renewed Helpdesk is tasked with promoting socially responsible IP practices — including time-limited, royalty-free licensing of EU-funded research in areas of societal concern — and, more strikingly, with helping to identify critical IP created through public funding and prevent it from leaving the EU without control and guarantees.
Read that last clause twice. The EU is starting to treat publicly funded IP as a strategic asset with, effectively, an export dimension. If your commercialisation plan for EU-funded research involves a non-EU acquirer or licensee, expect more scrutiny of that route over the next few years, not less.
2. Funding for university tech transfer — the call is open now
As we argued multiple times, brilliant research, premature disclosure, opaque licensing terms, and a Technology Transfer Office negotiation that outlasts the founding team’s enthusiasm can cripple commercialisation efforts (and interest).
The Commission has noticed. Call HORIZON-EIE-2026-03-CONNECT-01 (“From lab to market”) opened on 9 June 2026 and closes on 22 September 2026, funding around five projects at roughly €1 million each. Consortia need at least three universities or research organisations from three Member States or Associated Countries.
The deliverables are exactly what the sector needs: startup-friendly licensing templates, standardised IP transfer tools, and — crucially — benefit-sharing models that actually give researchers a reason to commercialise.
Let’s be honest about the money, though. Anyone who is — or has been — an academic knows that €1 million, split across a consortium of three or more institutions, barely scratches the surface. In practice, it buys more hours for your own research (which means teaching relief, if you hold a tenured position), a post-doc, and some research support. It’s certainly a promotion case if you come from a smaller university; nothing much at all if you come from an established institution. But it is a start — and it opens the door to the challenging, complex world of Horizon funding, which is worth far more than this first grant. So it is worth looking into.
These pilots feed into a planned EU-wide blueprint for licensing, royalty- and revenue-sharing and equity participation for academic institutions and their inventors — part of the “Lab to Unicorn” initiative. If it lands, it could do for spin-out terms what the UK’s USIT guides did after the 2023 spin-out review: give founders a benchmark to negotiate against. Watch this one.
3. Additional Funding
The 2027 calls under the same programme signal where the next funding waves will come from: an €18 million Startup Europe action prioritising deep tech companies in “moderate and emerging innovator” regions and women-led startups; a pre-commercial procurement action opening public-sector contracts to startups; and — genuinely novel — a €6 million action to pull philanthropic foundations into university commercialisation, backed by a Commission-funded mapping of Europe’s 33,000+ philanthropic organisations. Europe’s foundations sit on an estimated €567 billion in assets. Someone in Brussels has done the maths.
The Insights (honest) Analysis
€18 million is meant to be spread across nine projects and an entire continent’s startup ecosystem. Hold it against the scale of the problem it names: the Commission’s own study, presented at the European Parliament in late 2025, found that startups with at least one female founder attracted just 12% of European venture capital between 2020 and 2025 — and that only 16% of the general partners deciding where the money goes are women, managing a mere 9% of assets. All-female founding teams have been stuck around 6% of the market with stagnant capital for nearly a decade, and one estimate puts the cost of the deep tech gender gap at nearly €199 billion in lost European growth. Against a €199 billion problem, €18 million is not a strategy.
The philanthropic action fares no better under arithmetic. Six million euros to “engage” a sector of 33,000 organisations with €50 billion in annual expenditure and €567 billion in assets is roughly 0.01% of one year’s philanthropic spend.
However, it is important to bear in mind that these are coordination and support actions, not investment vehicles. They fund networks, matchmaking, and visibility. The theory is leverage, not volume; the actual capital is supposed to come from the EIC, InvestEU, programmes like Women TechEU’s non-dilutive grants, and the private market these actions are meant to unlock. However, signalling is not funding, and there is something uncomfortable about a work programme that names women-led startups as a priority and then prices that priority at less than a single Series A round. If it matters enough to write into the programme, it matters enough to fund at a scale someone might actually notice.
The uncomfortable truth: the environmental issue
Being analytical and writing insights means naming costs. Two of them deserve more than a footnote, and both will be explored in detail in other posts.
The environmental ledger is real, and it is not evenly distributed
Each gigafactory will consume roughly as much electricity as a medium-sized town, at a moment when European industry already pays around double the electricity price of its US competitors — a gap that any further geopolitical tension on energy markets will widen, not close.
Data centres need water: the EU plans to at least triple its data-centre capacity in the next five to seven years, and some of that expansion is heading into regions that are already water-stressed and getting hotter every year. In Aragón — one of Spain’s driest regions — hyperscale data centre projects are advancing while farmers watch the same reservoirs. The Netherlands and Ireland, meanwhile, have effectively stopped approving new data centres over grid and environmental concerns. Italy banned ground-mounted solar panels on agricultural land to prevent desertification while positioning itself as Europe’s fastest-growing data centre market.
Europe has been vocal — and rightly so — about climate change, and it has asked real sacrifices of the industries that built its economy, agriculture first among them. If those same governments now wave through the most energy- and water-hungry infrastructure of the decade, a credibility problem emerges.
Environmental and health disasters due to data centres are well known thanks to the infamous xAI’s Colossus supercomputer. The EU’s permitting and environmental rules are stronger than Tennessee’s — but rules only protect people if the pressure to “move fast on AI” doesn’t erode them.
And then there’s the question of who this is actually for
The financing model is deliberate: the EU covers up to 17% of a gigafactory’s computing infrastructure, Member States match it, and private capital supplies the rest. Brussels has set conditions — majority ownership must be European, high-risk vendors are excluded — but follow the gravity. The chips are Nvidia and AMD. The first gigafactory-scale AI facilities on European soil are being built commercially by Microsoft and its partners in Portugal, ahead of anything EU-funded. When a programme must “crowd in” tens of billions in private capital to work, the entities with tens of billions to deploy tend to write the terms.
Which brings us to the question: where is the opportunity for the SMEs and startups genuinely investing in innovation? On paper, the answer is generous — the AI Factories are explicitly designed to give startups, researchers and SMEs pay-as-you-go access to compute they could never afford alone. That is a real and underappreciated offer, and European founders should use it. But the risk is equally real: that public money de-risks infrastructure whose demand, pricing and priority access end up shaped by the tech giants, while European SMEs may get the leftover cycles and, through taxes, the possible invoice.
Europe is right to build. It is not yet right about how, where, and for whom — and those three questions deserve as much scrutiny.
Key Takeaways
The EU has decided that the answer to its commercialisation gap is not just regulation (its traditional export) but infrastructure, funding and standardised IP plumbing — gigafactories at the top of the stack, licensing blueprints at the bottom.
For SMEs and academics, the practical takeaway is the one our report keeps hammering: the support exists, but it rewards those who look. An SME paying full price for basic IP protection in 2026 — or a spin-out founder accepting whatever licence terms land on the table — usually hasn’t looked.
This post is general information, not legal advice. Consult a qualified IP professional before acting on anything here.
