The AI boom is rapidly reshaping global investment dynamics, but it is also raising concerns about opaque ‘black box’ strategies, heavy energy and water demands, and concentration risks.
That was the theme of a recently held roundtable in London by data provider Morningstar. Moderated by Morningstar senior editor Ollie Smith, the panel brought together Hortense Bioy, head of sustainable investing research, Morningstar Sustainalytics; Kenneth Lamont, principal for manager research, Morningstar; Nalin Patel, director of research, Emea private capital, PitchBook and Daniel Hayden, analyst for manager research, Morningstar.
EU-domiciled AI and big data funds reached a record $27 billion in the first half of 2025. However, Kenneth Lamont cautioned against assuming these trends will continue on a smooth path. “Thematic trends won’t necessarily follow linear trajectories,” he warned, noting that most AI and big data funds are still heavily concentrated in a handful of US tech giants. “Nine out of 10 AI and big data funds hold Nvidia,” Lamont said, highlighting the outsized exposure to the Magnificent 7 stocks ( Alphabet, Amazon, Apple, Meta Platforms, Microsoft, Nvidia and Tesla)
He added that AI comes with structural risks, including “massive energy and water consumption, regulatory uncertainty and potential geopolitical tensions.”
For fund selectors, things are getting more complicated. Lamont said: “You’ve got managers now using AI, whether it’s part of their process or, in some cases, just the main process.” With “AI” often used as a catch-all term, strategies can be hard to assess. “They can be black box,” he said, emphasising the need for transparency and explainability.
“In a deglobalising world, the big question is whether today’s US tech giants will stay on top,” Lamont added. “If you don’t understand it, don’t invest in it.” As a case in point, he referenced Nvidia’s sharp price drop in early 2025, a move triggered by changing valuation assumptions following the emergence of Chinese startup DeepSeek.
Nalin Patel pointed out that AI-related companies accounted for 50% of global VC-backed deal value and 30% of deal volume in H1 2025. In Europe alone, AI attracted €10.1 billion in venture capital and €4.2 billion in private equity over the same period. “AI now makes up a third of European VC deal value,” he said, though life sciences and fintech continue to hold their ground.
Patel stressed that the AI boom is no longer limited to just the startups building the core technology. “It’s not just about the core AI companies anymore,” he said. “We’re seeing a primary and secondary wave of impact.” That includes growing investment in energy-intensive data centres and semiconductor development — all critical to powering AI’s continued expansion.
“The public sector is moving too. Look at the EU’s AI Act. It’s a signal that regulation is catching up,” Patel said. On top of that, he pointed to new funding initiatives for AI-focused startups as a sign that governments are keen to shape, as well as support, the accelerating transition.
WATCH NOW: AI and the Future of Sustainable Investment
Daniel Haydon turned the spotlight on private markets. He welcomed their increased accessibility but raised concerns over lingering transparency issues. “Investors need to be careful not to get swept up in the sugar rush, especially with secondaries being bought at a discount and marked up straight away,” he said. With the secondaries market expected to hit $700 billion by 2028 and GP-led deals nearing $70 billion, he warned that such short-term bumps can distort underlying performance.
Haydon also flagged the complexity of fee structures in newer investment vehicles like interval funds. “Fees can range from 2% to nearly 8%,” he said. “At 8%, over 15 years, you’re paying close to your original investment in fees — that’s a big ask.” As more investors turn to products like LTAFs and Eltifs, Haydon questioned whether legacy equity risk models still apply, or whether “more dynamic, AI-informed approaches” are now needed.
“Clean energy is no longer just about combating climate change. It’s now a strategic enabler of the AI economy,” said Hortense Bioy. AI and clean energy are converging, driving the next wave of innovation and investment. Data centres already use 1.5% of global electricity, and by 2030, that’s set to double ( International Energy Agency), matching Japan’s current power consumption. In response, major tech firms including Microsoft, Google, and Amazon are ramping up their investments in solar, wind and even nuclear.
While AI stocks have soared, Bioy pointed out that “clean energy funds have struggled, with many showing negative returns over the past three years.” In contrast, private markets are delivering more attractive and stable returns. Bioy cited “Brookfield Renewable’s ~22% IRR since 2020” as an example of how private capital is capturing value through renewable infrastructure.
“Data centres are under pressure from all sides, be it energy, water or land. Innovation can’t be optional. The challenge is that the AI boom must be matched with responsible growth and sustainable practices,” added Bioy.
Finally, the panellists also touched upon the concern the UK has fallen behind in AI infrastructure, particularly in hardware and data centres. While the UK excels at producing software talent and AI research, it appears to be less involved in the foundational layers of the AI value chain. According to the panellists, UK companies may still benefit by being end-users of AI, especially in sectors like retail, where firms like Tesco are adopting AI to improve margins and efficiency. In that sense, even without owning the infrastructure, the UK could remain competitive by applying AI smartly across industries.
