The AI opportunity is a transformative technology adoption cycle which is in its early innings – a critical part of its adoption cycle. It’s a game changer for European business and will be a tremendous driver of economic growth and compelling investment returns in Europe. If European investors do not act now, the continent and its innovative companies will lose out, and compelling investment opportunities will be foregone. This is why private capital needs to act. If providers of private capital such as pension funds took minimal risk now, they would not only reap rewards in the future, but also boost the European economy.
But isn’t the AI game just for the US, with China nipping at its heels, and anyway isn’t it a bubble waiting to burst, I hear you ask?
Let’s go a little deeper and look at the AI tech stack. The foundation is semiconductors – for compute power, memory, networking, and storage required for AI training and inference – integrated into high-performance servers for AI data centres. Think Nvidia, Microsoft and Amazon Web Services. Then there are the models and operations, which includes foundational model development. Think OpenAI and Google. All of these companies are American. In fact, AI investments account for at least 40% of US GDP growth this year (FT: America is now one big bet on AI). AI data centre capital expenditure these days is approaching the astonishing level of 2% of US GDP. So yes, in this part of the AI tech stack the US leads globally. It does this by American entrepreneurs taking advantage of the massive pool of investment capital in the US. This is not where Europe will compete.
But the US has disadvantages too. The valuation bubble, which is a US phenomenon (BBC: America’s top banker sounds warning on US stock market fall). Deep political division. And plummeting levels of immigration – long the twin rocket fuel, alongside capital, for US entrepreneurship.
However, there’s another level of the AI tech stack, which builds on the layers where the US leads – the application layer.
This layer of the AI tech stack is now finding its way into mainstream business processes, signifying that the hype phase has passed and a generational technology shift has started. Its advent marks a lasting and profound change in the way whole industries operate. The space, or more accurately the spaces as it’s spawning multiple independent business lines and companies, is wide open and it’s where an enormous range of untapped opportunities lie.
- Imagine legal services which work speedily and efficiently, powered by specialist AI applications doing basic tasks working with and for qualified human lawyers who give legal advice. Anybody who has bought or sold a home knows how much time skilled lawyers spend on mundane tasks which could be done much better by AI.
- Imagine medical services where healthcare professionals spend all of their time caring for patients working with specialist AI applications which take care of the many basic, low skill tasks which take up their time today (FT: The transformative potential of AI in healthcare).
- Imagine students’ homework marked by AI, rapidly, and with bespoke comments and suggestions, freeing up the teacher’s time to prepare lessons and actually teach.
- Imagine the armies of people in finance company back offices, who today “stare and compare”, validating forms of data completed by their company’s customers, being replaced by AI. The machines will undertake these laborious tasks with far fewer errors and in a fraction of the time, freeing up the people to do higher added value work like decision making.
- Imagine vast libraries of video – we’re talking about millions of hours of sport, entertainment and other content types – which are understood and indexed by AI, allowing for example football clubs to discover, repurpose, share and monetise their footage.
These AI applications, and hundreds of others like them, are emerging today and are ripe for venture capital investment. They have the potential to inject rocket fuel into European industry, boosting the sectors where Europe holds its own and igniting the sectors which are already world class to extend their global leadership. Sectors like legal, pharmaceuticals, education, financial services and creative content in the UK. Companies are now beginning to adopt them and there are many years of adoption runway ahead for innovative businesses to grow into (Oxx: From Invention to Innovation: The Real Pace of Agentic AI Adoption).
So given the opportunity and the willing, world class customers, what holds Europe back?
It’s not entrepreneurship – as a venture investor over 30 years I’ve seen raw entrepreneurial talent, drive and hunger develop in Europe such that today it holds its own compared with the US. Furthermore, there’s now a generation of entrepreneurs who have learned their trade in European “founder factories” – organisations whose alumni have gone on to establish successful startups – like Revolut, Klarna, Wise, Spotify, Skype and Criteo.
There’s also plenty of talent in the ranks. Europe boasts four of the top ten Computer Science universities in the world (the UK alone has the top two) and 40% more developers than the US. It’s no wonder that Eric Schmidt, who when he was CEO of Google hired 10,000 people here, confirmed recently that Europe does not have a people problem.
What holds Europe back is access to capital.
This year US start-ups and scale-ups are taking in 62% of the world’s venture capital, whereas European companies are taking 15% of it (Dealroom: The State of Global VC). Europe has less than a quarter of the venture capital investment of the US, despite having twice its population and roughly equal GDP.
If this continues then what are its consequences for application layer AI companies? They’re very simple. American scale-ups will take on European markets, uncontested. Nature abhors a vacuum. The enormous AI adoption cycle will be another missed opportunity for Europe.
In the US much venture capital ultimately comes from investments by pension funds into venture capital managers. In fact some 10-15% of pension fund assets there are invested into private equity and venture capital combined. By contrast in Europe a mere 0.12% of pension fund assets are invested into venture capital (Pensions for Purpose: Venture & growth capital in Europe – mapping pension funds’ attitudes). Think about that for a moment. This means that of every €800 we’re investing for our retirement, just €1 is being invested into the technologies and products of the future. So the opportunity is being missed not just for the European economy today but also for the European pensioners of tomorrow.
The good news is that there are various initiatives around Europe to address this issue. For example the UK’s Mansion House Accord, France’s Tibi initiative and Germany’s WIN Initiative. The bad news is that they’re developing towards fruition at the pace of the pension sector, whereas AI is developing at lightning speed.
Having invested through multiple cycles – the dotcom build up, boom and bust, the credit crunch, the ZIRP era – and seen the breadth of the industry from my time with the BVCA, I’m cursed with a premonition of what will happen. The European pensions sector will not invest much at all in venture capital for the next few years. It will miss the most attractive stage of the AI adoption cycle when the best investment returns are there to be made. Then a switch will flick and the capital will flood in. This glut will cause a bubble, which will burst a few years later, leading to negative returns, disappointed investors and the start of a new nuclear winter for European venture capital.
But it doesn’t have to be this way.
Private capital, whether from the pensions sector or beyond it, could take up the slack and invest through venture capital funds into the early stages of the AI application layer adoption cycle in Europe. They will be driving economic growth in Europe and capturing the attractive investment returns which are there to be made.
So, what should be done? If you’re a pension fund in Europe then you should start a pathfinder program. Dip your toe in the water with initial modest investments into a handful of venture funds with proven investment track records to get to know the sector, and to develop expertise in selecting venture capital managers. If you’re a non-pension fund private sector investor in Europe, then you should ask yourself whether you would rather have invested into the Internet sector in 1995 or in 2000. If the answer is 1995, then you should fill the void left by the current low level of pension fund investment in venture capital and get in early to the application layer AI adoption cycle. And if you’re a private sector investor from outside Europe then you should look closely at the European venture capital sector, comparing European VC funds with US VC funds in terms of track record as well as market opportunity, and draw your own conclusions.
The opportunity to invest in application layer AI in Europe through venture capital is compelling, and the time to do it is now. Private capital has the chance to unlock innovation, generate exceptional returns, and ensure that Europe captures its share of the AI revolution. Investing in European venture capital will be a win for investors, and a win for Europe.
