When it comes to artificial intelligence (AI), American hyperscale technology companies are getting a lot of attention. However, European companies may have an advantage when it comes to building applications for AI.
European ventures are at the forefront of developing applications that sit on top of large-scale language models (LLMs) and are driving growth in this area, said Cliff Marriott, co-head of the EMEA Technology, Media and Telecommunications Group, Global Banking & Markets.
“Europe is fighting back,” Marriott said. “European high-tech companies are leading the way in 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 in March this year. The mood was upbeat, especially for early-stage founders and investors.
The number of European unicorns (start-ups worth at least $1 billion) has more than tripled to 413 since 2016, according to data from London-based venture capital firm Atomico. And from 2025 to early 2026, approximately 30 new unicorns were minted.
Why software companies can benefit from AI
Symposium attendees also focused on the impact of AI on the software industry’s business models, a challenge that is causing uncertainty in public markets. That said, the conference was filled with fresh ideas about how the industry will adapt, said Joe Porter, global co-head of software investment banking in the Technology, Media and Telecom Group, Global Banking & Markets.
“Despite the headlines, incumbent software company CEOs have the trust of their customers for distribution, implementation, and long-term customer relationships to help them drive revenue with new AI-driven capabilities,” Porter says. “Many business leaders feel this is not just a threat, but a huge opportunity.”
We spoke to Marriott and Porter about Europe’s AI push, the impact of AI agents, and the rise of Europe’s burgeoning defense technology (DefTech) sector.
Given the challenges of AI, what is the outlook for the software industry?
Joe Porter: First, technology in general is deflationary, and AI has proven to be precisely deflationary at its most maximal level in terms of code writing and software development. The cost of developing code is decreasing, and at the same time the speed of development is accelerating. The presence of market forces in the software layer means increased competition.
There are several lines of battle that I would like to highlight. First, what is the value of writing code itself? Just because you can write a line of code doesn’t necessarily mean it’s a functional product, built specifically for enterprise IT environments. That doesn’t mean you know what to prompt or model, or what software capabilities are needed to connect to different datasets or other systems. And that certainly doesn’t mean that software customers will buy this in every case.
And the second battle is software pricing. How do people pay for the value of agent- or outcome-based software? This is a central debate in a world of AI and AI-driven software. The question is who will capture that value, compared to new competitors that may have AI technology advantages today, versus existing cloud software players with customer relationships and large-scale delivery capabilities. What is the value of software distribution and customer relationships today?
I call this client trust, but how much is it worth in the world of AI compared to the creation of the code or product itself?
These sound like basic questions.
Joe Porter: I don’t think it’s an exaggeration to say that the fundamentals and economics of software are being questioned or confused.
What advantages does the European technology community have when it comes to AI?
Cliff Marriott: If you go back over the last 10 to 20 years, the places where Europe has excelled are also the places where Europe should have excelled. Consider fintech. Scaling up in Europe required dealing with multiple currencies, regulations, and borders. In the US, you can basically create one product that works relatively seamlessly, but when it comes to global expansion, it may not be as adaptable. Europe is doing very well in fintech.
In the future, the most active area will be the creation of AI applications for enterprises, or the application layer. Although some great companies are in the fight, it is not yet clear whether Europe will emerge victorious in the large-scale language modeling (LLM) space.
In the AI application layer, there are many companies in different parts of the continent that are doing well and are scaling up. They have received great response from customers and have started raising capital at attractive valuations. There are question marks over whether they can sustain their success, whether they can scale enough to be a winner, and how LLMs will respond. Will it go vertically?
AI agents have been a hot topic this year. How disruptive will they be to businesses?
Joe Porter: If today’s software makes humans more efficient or achieves some degree of automation, agent AI will evolve applications to complete tasks and produce results. At the symposium, I chaired a panel that included executives from various walks of life. We had developers, sales and marketing executives, and people from companies that specialized in documentation, or what I call knowledge work.
The programmer believes that in the future there will be as many engineers as there are today working with AI agents to focus on higher-level IT strategy and development issues. Salespeople were less confident that there would be as many salespeople, given that many tasks will be automated in the future.
What was interesting to me was that they all said there would be humans involved, not just agents. The debate is where agents stop and humans begin. This will get us all to the point where we can all be much more efficient because we are letting the actual software complete the tasks instead of just doing it ourselves.
Cliff Marriott: A simple explanation of Joe’s point is that people in organizations become more productive over time. You can imagine a scenario where you have an incredibly large corporate barbell at one end, and a single individual entity at the other end, and by extension, it’s just individuals and their agents who are creating software and their businesses. I may be oversimplifying here, but imagine a scenario where a company is founded and evolves.
What is driving the growth of European defense technology startups?
Cliff Marriott: There has been a theme over the past 30 years that Europe has not sufficiently developed its defense infrastructure. The way we manage war and defense is not with big artillery or tanks, but with new technology and drone systems. And it’s not just airborne drones. It’s also on the sea. Seven defense companies participated in the conference, and there is a lot of activity in this area as companies expand. Governments and the Department of Defense are rearming and preparing for a new world that requires more ground capabilities.
Given these trends, will European IPOs and other types of exits increase?
Cliff Marriott: Many people focus on IPOs as a method of exit, but in reality, most companies exit through M&A, and I think M&A will become very active in the short to medium term.
What’s interesting is how much time we spend with private companies that decide to remain private for an extended period of time. This is not because they cannot go public, but rather because they can achieve many objectives as a private company. We also raise a number of privately placed bonds for companies considering investing in AI.
Joe Porter: I think we will continue to see large strategic software companies buying new-age AI technologies through acquisitions. they are hunting.
Unlike temporary economic shocks or macro events, this technological change is evolutionary. For large companies, small businesses, and companies in between, standing still is not a strategic option.
This article is provided for educational purposes only. The information contained in this article does not constitute a recommendation by any Goldman Sachs entity to any recipient, and Goldman Sachs does not provide financial, economic, legal, investment, accounting, or tax advice through this article or any recipient thereof. Neither Goldman Sachs nor any of its affiliates makes any representations or warranties, express or implied, as to the accuracy or completeness of any statements or information contained in this article, and any liability therefor (including for any direct, indirect, or consequential loss or damage) is expressly disclaimed.
