EU Says It Funds Deep Tech Led by Women. The Data Says Otherwise.

TL;DR: EU Says It Funds Deep Tech Led by Women. The Data Says Otherwise.

Europe has pledged billions to support deep tech innovation and gender equity simultaneously. In practice, female-led deep tech companies receive just 11% of sector funding, the EIC Accelerator rejects over 93% of all applicants, and startups genuinely building hard technology from scratch face structural bias toward safe, credentialed, consortium-backed applicants. The funding architecture is not broken. It is working exactly as designed, and that design does not prioritise you.

💡 Check out this opinion piece on EU funding for female entrepreneurs..

I am the CEO and co-founder of CADChain. We build blockchain-based intellectual property protection software for CAD files and 3D models. That puts us squarely in deep tech: hardware-adjacent, IP-intensive, technically complex, slow to build, and far outside the comfort zone of most evaluators who have never shipped a product.

I have also spent years applying for EU funding across multiple programmes, writing about the dysfunction in that system for Sifted, and watching EU-funded projects ask female deep tech founders to work for free while sitting on millions in public money. If you have not read those pieces yet, the link is at the top of this article.

This one is specifically for founders building hard technology. Because the situation in deep tech funding is worse than the general picture, and the specific ways it is worse matter to anyone deciding whether to spend months on a grant application.

Here is what the data actually says.

The Deep Tech Gender Gap Is the Worst Gap in the Room

Start with the headline number from the 2026 European Deep Tech Report. Female-founded companies attracted only 14% of European deep tech VC funding in 2025, with no meaningful improvement over eight years.

Eight years. Not a trend. A fixture.

The EIT's 2024 study on deep tech funding sharpened that picture further. Female-led deep tech startups receive just 11.4% of total sector funding. At the seed stage specifically, women-led companies secure 15% of available capital. At early and late-stage VC, that drops to 11%.

Let that progression sit for a moment. As a female-led deep tech company matures, its share of available funding shrinks. The later-stage capital that enables real scale flows overwhelmingly to male-led teams. A 2025 European Commission document confirmed the same pattern in blunter terms: in tech startup capital investment, all-female founding teams receive 3% of the total.

Three percent.

And this is occurring inside a sector that the EU publicly identifies as a strategic priority for European sovereignty, competitiveness, and the energy transition.

The EIC Accelerator: Europe's Flagship Deep Tech Programme

The European Innovation Council Accelerator is the EU's primary instrument for deep tech funding. It offers up to 2.5 million euros in non-dilutive grants and up to 10 million euros in equity per company, under the Horizon Europe framework. The EIC budget for 2025 reached 1.4 billion euros, up nearly 200 million from 2024.

Those are impressive numbers. Here is what the application data looks like.

The cumulative full-application-to-win rate across five years of the EIC Accelerator sits at 7%. That is across roughly 12,000 full applications submitted between 2021 and 2025, resulting in around 700 funded companies. In the October 2024 cutoff, the total success rate including Step 1 was approximately 5.9%.

To reach the full application stage, a startup first needs three out of four remote evaluators to award a GO at the short application stage. The rate for short applications passing through has dropped from 65% in the early years to 35% in 2024, following a rule change requiring three GOs instead of two.

On top of that, a rule introduced in 2024 permanently bars a company from reapplying under Horizon Europe after three NO GO decisions across any stage. For a startup building something genuinely new, where the first attempt inevitably reveals gaps, this is a structure that punishes the learning curve.

For female-led deep tech startups specifically, these odds compound with the baseline gender funding gap. Twenty-one percent of funded EIC Accelerator companies in 2024 were female-led, rising to approximately 30% in 2025. Progress, yes. But female-led deep tech startups make up a far higher percentage of applicants than their 21% funding share reflects, and the EIC itself acknowledges the gap by specifically welcoming female CEOs in its programme language.

Welcoming language is not the same as structural change.

What "Deep Tech" Means When Europe Evaluates It

At CADChain, we secure intellectual property for CAD files and 3D models using blockchain. Engineers and manufacturers use CAD software daily, and the IP embedded in those files represents enormous economic value that is almost entirely unprotected. A design file sent to a supplier, a contractor, a manufacturer overseas, carries no built-in protection. Once it leaves your system, your ownership claim exists only on paper.

We were building a solution to a real and measurable problem. We were doing it in deep tech, with blockchain as the technical layer, targeting an industrial sector that Europe claims to prioritise for its strategic autonomy.

We got some funding but it turned into a bureaucratic nightmare.

The feedback patterns I encountered across applications, and which I described in my Sifted piece on EU funding nightmares, are consistent with a system that defines deep tech risk in a very specific way. Evaluators reward proofs of concept that have already been validated. They reward teams that have already received funding. They reward projects that resemble previous projects.

In other words, they reward incumbency. And for a startup building something that does not yet have a category name, incumbency is precisely what you lack.

The EIC Pathfinder, which specifically targets science-to-technology breakthroughs at TRL 1 to 4, explicitly states in its call documentation that projects must have a long-term vision for a transformative technology and a high-risk, high-gain methodology. It also requires intellectual property protection and exploitation planning.

That is the correct language. The question is whether evaluators who assess those proposals have ever tried to build something genuinely high-risk. Based on the feedback patterns I have observed over years, the answer is often no.

The IP Problem That Nobody Funds

CADChain sits at an intersection that the EU says it cares about: intellectual property protection, manufacturing, industrial digitisation, deep tech, and female-led startups.

In practice, we have found that IP protection as a product, rather than IP protection as a credential, is not a category that EU grant programmes handle well. The EU's IP Helpdesk, the EUIPO, and the EPO all provide guidance and support for startups looking to protect their own IP. That infrastructure is genuinely useful.

What does not exist, at any meaningful scale, is EU funding for startups that are building the tools to help other companies protect their IP in real-time, programmatically, at the file level. That is a harder sell than a medical device or a synthetic biology platform, because the problem it solves is invisible until it is too late.

More than 193,000 patents were registered through the European Patent Office in 2023 alone, according to Euro-Funding data. European deep tech leads globally in IP protection infrastructure. And yet the companies building next-generation IP tools for the manufacturing sector sit outside the standard funding frameworks, because their category does not map cleanly to health, energy, or climate, the three dominant verticals in EU deep tech funding.

The 2026 EIC work programme allocates 634 million euros to the EIC Accelerator Open Call, 120 million to Challenges focused on autonomous robotics, climate crops, waste conversion, and medical diagnostics, and 300 million to the STEP Scale-Up scheme for companies ready to grow to 50 to 150 million in total investment with private co-funding.

Industrial IP protection tools for manufacturing SMEs do not appear in any of those categories as a defined priority. That does not mean they are ineligible. It means the path to funding is longer, harder to narrate, and requires far more evaluator education than a startup in a recognised vertical.

The Consortium Lock-Out, Repeated

For larger EU grants, the consortium model applies here exactly as it applies in the female entrepreneurship space I described in my other pieces. Deep tech startups without five years of programme experience cannot create a new consortium and cannot easily join an existing one.

The organisations that win large deep tech consortium grants are, in many cases, universities, established research centres, and large industrial companies with dedicated grant teams. They have the administrative infrastructure to absorb the reporting burden. A startup with six employees does not.

The EIC Accelerator is at least structured as a single-entity instrument, which removes the consortium requirement and makes it genuinely accessible to small teams. That is a real advantage. But the 7% success rate means that for every startup that succeeds, thirteen do not, and each of those thirteen has spent months on an application that could have gone toward building product.

The question every deep tech founder should ask is: what is the opportunity cost of a grant application cycle, and does the expected value justify it?

For an early-stage company where time and engineering focus are the scarcest resources, the answer is often no. Especially when the architecture systematically disadvantages new entrants, female-led teams, and startups in categories that evaluators have not seen before.

What CADChain Did Instead

CADChain's core IP protection tools exist. Boris for Inventor is in beta. BlendedBoris for Blender is in alpha. The SolidWorks integration is running. The 3D printing proposition is in stealth.

We built on lean budgets, with a small team, without the runway that a successful EIC Accelerator application would have provided. The IP protection problem we are solving is real, the market is real, and European manufacturing companies need this whether EU grant programmes have a category for it or not.

And we have received emails, like clockwork, from EU-funded projects asking us to contribute our expertise for free. Projects with millions in Horizon Europe funding, asking a female-led deep tech company to provide technical workshops without compensation. I documented the full mechanics of how that works in my piece for MeanCEO, which is linked at the bottom of this article.

The irony is complete. The system that would not fund us now wants to extract value from us to justify its own existence.

The One Data Point That Offers Actual Hope

The EIC's focused effort on gender equity within its own programmes produced a real result: female founder participation rose from 8% in 2020 to 30% in 2024 inside EIC-supported companies, according to EISMEA data presented in Brussels in late 2025.

That is what intentional structural intervention looks like. Not workshops. Not cascade funding with mandatory consultant cuts. A targeted programme, run inside the main funding instrument, with a measurable target and public accountability for the outcome.

The Women TechEU programme, which provides 75,000 euros per company in non-dilutive grants to early-stage female-led deep tech startups, is another example of a focused instrument that actually reaches founders rather than circling back to consortium overhead.

These are small numbers relative to the EIC's total budget. But they are the right shape of intervention.

Next Steps for Deep Tech Founders

If you are building hard technology and evaluating whether EU funding is worth pursuing, here is the practical framework.

For the EIC Accelerator, the 7% success rate means this is a real lottery that also requires months of preparation. It is worth one serious attempt if your technology is at TRL 4 or above, your IP position is solid, and you have a clear commercialisation narrative. The new 20-page full proposal format for 2026 reduces the paperwork burden. The three-strikes-and-out rule means you need to treat each attempt as if it is your last.

For Women TechEU, if you are pre-MVP and female-led, the 75,000 euro non-dilutive grant with no mandatory consultant cut is one of the cleaner instruments available. Apply specifically for this before anything else.

For any cascade funding programme that approaches you with a collaboration offer: ask what percentage of the total project budget reaches third parties, and ask whether your time will be compensated. If the answer to the second question is no, you are being asked to subsidise someone else's grant. Decline.

And if you are building IP protection tools, manufacturing technology, or industrial deep tech that does not fit cleanly into health, climate, or energy, know that you are operating outside the standard EU funding categories by design. Build the company anyway. The problem you are solving does not care whether it has a grant programme attached to it.