Hyperscale technology companies in the US have captured much of the attention when it comes to artificial intelligence (AI). However European companies may have an edge when it comes to building applications for AI. 

European ventures are at the forefront of the development of applications that sit on top of large language models, or LLMs, that are driving the growth of the sector, says Clif Marriott, the co-head of the Technology, Media, and Telecommunications Group in EMEA in Global Banking & Markets.

“Europe is battling back,” Marriott says. “One could say that European tech enterprises are leading in terms of developing AI application layer companies.”

This was a key takeaway from the 11th Disruptive Technology Symposium, which brought together more than 1,300 investors and 72 startups in London this March. The mood was buoyant, especially for the early-stage founders and investors.

The number of European unicorns—startups worth at least $1 billion—have more than tripled, to 413, since 2016, according to data from Atomico, a London-based venture capital firm. And almost three dozen new unicorns were minted in 2025 and early 2026.

Why software companies will benefit from AI

Symposium attendees were also focused on AI’s impact on the software industry business model, a challenge that has caused jitters in the public market. That said, the conference buzzed with fresh ideas about how the industry will adapt, says Joe Porter, the global co-head of software investment banking in the Technology, Media, and Telecom Group in Global Banking & Markets.

“Despite the headlines, the CEOs of incumbent software companies have client trust with distribution, implementation, and long-standing customer relationships that will help them monetize with new features and functionality that are AI driven,” Porter says. “There are a lot of company leaders that feel this is a massive opportunity, not just a threat.”

We spoke with Marriott and Porter about Europe’s push in AI, the impact of AI agents, and the rise of Europe’s burgeoning defense technology, or DefTech, sector.

Given the challenge of AI, what is the outlook for the software industry?

Joe Porter: First, technology is generally deflationary, and AI is proving to be just that at the most maximum level, in terms of code creation and software development. The costs of developing code are coming down at the same time the speed of development is accelerating. When you have those market forces in the software layer you are going to see an increased amount of competition.

There are a couple of battle lines I would highlight. One, what is the value of code creation in and of itself? Just because code lines can be created doesn’t mean it is necessarily a functional product and purpose-built for enterprise IT environments. It doesn’t mean that we know what to prompt or model, or what features of software are required to tie into various datasets and other systems. And it certainly doesn’t mean that a customer of software would buy this in all cases.

And the second battleline is the pricing of software. How do people pay for the value of agentic or outcome-based software? This is the core debate in a world where there is AI and AI-driven software. There are questions of who is going to capture that value across new competitors that may have an AI technical advantage today versus incumbent cloud software players who have client relationships and large distribution capabilities. What is the value of software distribution and client relationships today?

I call this client trust, and how much is it worth in an AI world relative to the code and product creation itself?

Those sound like fundamental questions.

Joe Porter: I think it’s fair to say the fundamentals and economics of software are being questioned or disrupted.

What advantage does Europe’s tech community have when it comes to AI?

Clif Marriott: If you go back over the last 10 to 20 years, where Europe excelled is where it needed to excel. Take fintech. To scale in Europe, you had to deal with multiple currencies, regulations, and borders. In the US, one can basically create one product that works relatively seamlessly, but it may not be as adaptable when it comes to going global. Europe has done really well with fintech.

Going forward, we see the most activity and focus on the creation of AI applications, or the application layer, for companies. Although there are some great companies battling, it is not yet clear Europe will have a winner in the large language model, or LLM, space.

In the AI application layer, there are a bunch of companies in different places across the continent that are performing well and scaling. They are having great resonance with customers and are starting to raise capital at attractive valuations. There is a question mark on whether they will be able to sustain their success, will they scale enough to be winners, and how will the LLMs react. Will they go vertical?

AI agents have become a big story this year. How disruptive will they be for businesses?

Joe Porter: If software today makes a human more efficient or helps achieve some level of automation, then agentic AI will evolve applications, complete the task, and produce outcomes. At the symposium, I chaired a panel with executives from different points of view—we had a developer, a sales and marketing exec, and someone from an enterprise that specializes in document creation or what I call knowledge work.

The coder believes there will be as many engineers in the future as today to work alongside AI agents and focus on higher level IT strategy and development problems. The salesperson was less confident that there’s going to be just as many salespeople given many of the tasks will be automated going forward.

What was interesting to me is that all of them said there is going to be a human in the loop, it won’t just be agents. The debate is where do the agents stop and the humans begin. This is going to get to a spot where we are all just going to be a lot more efficient because we will have actual software completing the task rather than just completing it ourselves.

Clif Marriott: Just to pick up on Joe’s point, you can see people at organizations becoming more productive over time. I could imagine a scenario where we end up with barbells of incredibly large enterprises on one end and then one-person entities on the other, which can also be scaled but is just a person and their agents creating software and its business. Maybe I’m oversimplifying here but you can imagine this scenario of company formation and evolution.

What’s driving the growth of European defense technology startups?

Clif Marriott: There is this theme that Europe has not developed enough of its defense base over the last 30 years. The way warfare or defense is being managed is by new technology and drone systems rather than large artillery or tanks, etc. And it isn’t just airborne drones. It’s on the sea as well. We had seven defense companies at the conference and there is a lot of activity in the space with companies scaling. It’s governments and defense departments rearming and preparing for a new world where they need more on the ground capabilities.

Given these trends, will European IPOs and other types of exits pick up?

Clif Marriott: A lot of people like to focus on IPOs as the way you exit but the reality is that most companies exit through mergers and acquisitions, and I think M&A will be quite active in the near to medium term.

What’s interesting is how much time we are spending with private companies that are deciding to stay private for longer. That’s not because they can’t go public, but rather because they can achieve a lot of their objectives as private companies. We are also doing a lot of private debt raising for companies that are looking to invest in AI.

Joe Porter: I think we are going to continue to see large strategic software companies buy new age AI technology through acquisition. They are on the hunt.

Unlike an economic shock or a macro event that may be temporary, this technology shift is evolutionary. Standing still is not a strategic option for big companies, small companies, or anyone in between.

This article is being provided for educational purposes only. The information contained in this article does not constitute a recommendation from any Goldman Sachs entity to the recipient, and Goldman Sachs is not providing any financial, economic, legal, investment, accounting, or tax advice through this article or to its recipient. Neither Goldman Sachs nor any of its affiliates makes any representation or warranty, express or implied, as to the accuracy or completeness of the statements or any information contained in this article and any liability therefore (including in respect of direct, indirect, or consequential loss or damage) is expressly disclaimed.