European startups raised €21.45 billion ($24 billion) in the second quarter of this year, the region’s strongest quarter for venture funding in four years, according to Crunchbase. On the surface, capital is back.
Look one level down and the picture changes. 65% of that money went to 42 companies raising rounds of €90 million ($100 million) or more. Investors are also becoming more disciplined about where that capital goes. OMERS Ventures described its outlook for 2026 as a “return to discipline,” with deeper diligence around traction, retention, defensibility and resilience.
The bar for raising has gone up, but not for the reason most founders assume. It was never that investors stopped writing cheques. It is that they are choosing harder, and one of the places that scrutiny shows up is the moment they ask a founder about intellectual property.
Investors are not asking whether you have patents
When an investor asks about IP, they rarely mean it in the literal sense. “We’ve filed three applications” tells them almost nothing. They want to know whether the founder understands which part of the company is genuinely hard to copy, and whether the protection in place covers it.
Two things are happening here, though most founders only see one of them.
First, the cost of building software has fallen dramatically. In its 2026 outlook, OMERS Ventures argued that AI development tools are making software faster to build, while durable moats in pure software are becoming harder to establish and maintain. Software has not stopped being valuable; founders need a more precise answer to what makes their business defensible.
Second, the link between IP and funding outcomes investors care about has been in the data for years. A joint study by the EUIPO and the EPO found that startups with patent applications at seed or early-stage stages were associated with a 6.4 times higher likelihood of subsequent funding.
That does not mean filing a patent makes a company fundable. Investors are not rewarding the document. The filing can be a signal that the company has identified something differentiated enough to protect.
Put those together and the question changes. Instead of “Do you have IP?”, it becomes “Does your IP protect the part of this business a competitor cannot simply rebuild?”
A patent that does not map to the business is a cost, not an asset
The most common weakness is not the absence of patents, but patents filed for the wrong reasons. The pattern is familiar: claims drafted around how the product looked on filing day rather than the mechanism that makes it work. Six months later, the product has moved on, and the claims protect a feature that is no longer important to the company.
A strong answer sounds different, naming the specific technical mechanism the claims cover. It explains why the company filed where it did: a regional application because the commercial plan runs through several markets, or a national filing first with a Patent Co-operation Treaty (PCT) application to keep options open. And it shows clean ownership, with every contribution assigned to the company, including work done by contractors and by the founders themselves before incorporation.
That last one catches more companies than any other. A founder who wrote the core algorithm six months before the company existed often still owns it, and nobody has noticed until diligence.
It is worth being clear about the limits of this. A patent is not a business moat just on its own, and I don’t think anyone can say with confidence which forms of defensibility will hold as AI keeps compressing the time it takes to build.
Proprietary data, distribution, and speed of execution all matter, and some investors will weigh those factors as heavily as, or more heavily than, anything filed at a patent office. But that is an argument for a deliberate IP strategy, not against one. The point is not to patent everything. It is to understand what is strategically important enough to protect and make sure the protection follows the business.
The fundraise is the worst time to build an IP strategy
Most founders discover their answer to the IP question during diligence, the moment they have the least time and least leverage. The aforementioned EUIPO-EPO study shows how common this is: only 10% of startups raising seed rounds had filed for patents, rising to 28% at the early growth stage and 44% at late stage.
By the time diligence begins, however, an investor is not only looking at what has been filed. They are looking backwards at the decisions that produced the portfolio in front of them.
The reasons are the ones that hold back IP protection all across Europe: cost, complexity and time. An application drafted in a rush the week before a data room opens is rarely one that will hold up, and founders know it.
That is where the economics have started to shift. In AI-native patent services, the technology produces the invention disclosure, the prior art search and the first draft of the specification. A named attorney, qualified before the relevant patent office, then reviews the claims, refines them and takes responsibility for what is filed. That puts considered protection within reach early, when it can shape the business rather than describe it after the fact.
No one expects a founder to be a patent attorney. But by the time a company is raising serious capital, someone should have asked the harder questions: What is actually novel here, what would matter if a competitor copied it, and does the protection we have still match the business we are building?
Diligence is a good place to test an IP strategy. It is a bad place to discover you never had one.