Fundraising, AI and Europe’s Capital Problem

For years, venture capital has relied on familiar frameworks. Investors looked at recurring revenue, benchmarked SaaS multiples and built conviction around predictable growth, but now, AI is starting to expose the limits of that thinking.

This article is based on a recent conversation on Nothing Ventured between Aarish Shah and Ben Prade, Partner at Bullhound Capital (formerly GP Bullhound). Having spent close to a decade building the firm’s growth investment strategy across frontier technologies, Ben shared his perspective on why venture capital still struggles to price AI companies, what that means for founders raising capital, and how the conversation extends far beyond valuation into fundraising, deep tech and Europe’s position in the global technology race.

Many AI businesses don’t fit the subscription model that defined the last generation of software. Revenue increasingly comes from usage, often measured in tokens rather than seats or licences, which makes both pricing and valuation harder. A customer might spend heavily one month and then dramatically less the next, while costs can fluctuate too, particularly when companies rely on large language models or expensive compute. As a result, the relationship between revenue, margin and growth is far less straightforward than investors have grown used to.

That uncertainty means the tools used to value them haven’t caught up. The market still wants clean comparisons with SaaS because that’s the language investors know. AI companies are forcing a different conversation.

Some organisations are already spending astonishing amounts on AI simply to build capability. Token budgets disappear in months instead of years. In many cases, falling behind is seen as the greater risk. Companies are accepting short-term cost because they believe AI adoption is becoming a competitive requirement rather than an optimisation exercise.

That uncertainty carries into fundraising.

Founders often assume raising capital is primarily about explaining the technology. For deep tech companies, technical credibility matters, but it’s rarely enough. Investors need to understand why the technology matters, why this team is the one to commercialise it and why the timing is right.

Bullhound’s own hiring reflects that balance. The firm has moved towards recruiting PhDs and highly technical specialists who can properly assess complex science. Yet technical expertise alone doesn’t persuade investors. Scientific accuracy and commercial storytelling have to work together. The strongest founders don’t ignore the details. They build a narrative that helps investors understand why those details matter.

The challenge becomes even greater in deep tech because the capital requirements are larger. Companies developing quantum computing or advanced AI infrastructure often need funding rounds that extend well beyond what early-stage investors can support. That creates pressure to attract growth investors while much of the technical risk still exists. By that stage, belief isn’t enough. Investors want evidence through customers, intellectual property, commercial traction or proof that previous funding has been converted into meaningful progress.

This is where Europe’s structural problem begins to emerge.

Europe produces exceptional research, talented founders and globally relevant technology. What it consistently struggles to provide is enough late-stage capital to keep those companies independent as they scale.

Seed funding has improved considerably, but growth capital remains limited. As companies mature, many have little choice but to seek investment from US funds. That often brings access to larger pools of capital and bigger markets. It can also shift decision-making, talent and, eventually, headquarters away from Europe. The issue isn’t simply ownership. It’s whether Europe captures the long-term economic value created by its own innovation.

DeepMind is an example that still shapes conversations across European venture capital. Its acquisition demonstrated that Europe can produce world-class AI businesses. It also reinforced concerns that the region struggles to retain them long enough to realise their full value.

Governments are becoming increasingly important players in trying to close that gap.

Quantum computing illustrates the shift. Bullhound’s investment in Oxford Quantum Circuits came alongside substantial UK government backing. Public funding doesn’t remove commercial risk, but it changes the equation for private investors. Government support signals that certain technologies have strategic importance beyond financial returns.

That creates another layer of complexity.

Frontier technologies increasingly overlap with national security, defence and economic sovereignty. The source of capital starts to matter alongside the amount of capital available. Sovereign investors may impose restrictions on where technology is developed, how intellectual property is managed or who can participate in future funding rounds. As a result, fundraising becomes about more than valuation. Founders may find themselves making long-term strategic decisions about ownership, geography and control much earlier than previous generations of startups.

Europe faces a difficult balance. It wants deeper pools of private capital while also preserving strategic autonomy in technologies such as AI and quantum computing. Those objectives don’t always align. The more dependent companies become on overseas funding, the harder it becomes to keep decision-making local. Yet limiting access to global capital creates its own risks if promising businesses can’t scale.

There isn’t a simple solution. Larger specialist funds may help support capital-intensive sectors, but concentrating investment in narrow technologies increases risk for investors. Governments can fill some of the funding gap, though political cycles rarely match the long development timelines required for frontier technologies. The discussion points towards a need for more consistent, long-term thinking if Europe wants to compete globally rather than produce companies that eventually scale elsewhere.

For founders, the implications are becoming clearer.

Building a great product is only one part of the job. Capital strategy is becoming a competitive advantage in its own right. The way a company prices AI, tells its story and chooses its investors may shape its future just as much as the technology it builds.

As AI continues to reshape software, venture capital is being forced to rethink some of its oldest assumptions. The question is whether capital, policy and company building can evolve at the same pace as the technology itself.