# GIC bullish on AI but warns of capital concentration as ‘largest market risk’ > Source: > Published: 2026-07-23 21:00:00+00:00 # GIC bullish on AI but warns of capital concentration as ‘largest market risk’ It is also optimistic on renewable power for data centres, transmission infrastructure [SINGAPORE] Singapore sovereign wealth fund GIC is upbeat on the strategic opportunities that artificial intelligence offers, but cautioned that the amount of capital flowing into the sector poses a major concentration risk. “Globally, more and more capital is going into the AI ecosystem, which is... one highly correlated, concentrated risk factor,” said GIC group chief investment officer Bryan Yeo at a media briefing ahead of the report’s release on Friday (Jul 24). This is “the largest market risk factor out there – more so than any country, industry or sector”, he added. Yeo was responding to a question from *The Business Times* on the risk of an AI bubble and how GIC is managing its exposure. The fund has [bet big on AI giant Anthropic](https://www.businesstimes.com.sg/singapore/gic-invest-us1-5-billion-ai-start-anthropic-ft-reports) since 2025, most recently co-leading its US$65 billion fundraise in May. Other notable investments include Ramp, an AI-powered financial operations platform, and data management company Databricks. GIC assesses its aggregate portfolio exposure to AI, evaluating if the level is too high, too low or just right, said Yeo. “It’s not just the headline number; we also go two to three layers deeper – What is the composition of these (sources of) AI exposure we have, how correlated are they?” he noted. Despite the risks, GIC is still bullish. In the past few years, it has seen AI as “genuinely transformative”, said Yeo. “AI is one area (where) we have articulated that we would invest in, since 2022... Since then, we have furthered our conviction in this space as AI has reached a point of real-world impact,” he noted. “It is no longer a question of whether it can create economic value, but where and how fast.” In a survey of almost 200 North American companies in the first quarter, GIC found that more than half of the respondents ranked AI among their top three priorities. Within this group, one in five named it as their top priority. The companies also reported cost reductions from AI in areas such as finance, software development and customer service. Yeo expects more of such gains to come, but noted: “As AI scales, the ecosystem gets a lot more complex. It is also getting harder to pick winners.” GIC is also mindful that “near-term breakthroughs are not always indicators of long-term winners”, he added. ## Adopters as the next big leg-up To navigate this complexity, GIC classifies AI opportunities into three categories: enablers, monetisers and adopters. Enablers build the foundational infrastructure for AI, such as computing hardware layers and high-speed networks. GIC’s investments in this category include Anthropic and hyperscale player Vantage Data Centres. Monetisers develop AI-powered products and services, such as [San Francisco-based Databricks, which GIC backed](https://www.businesstimes.com.sg/startups-tech/ai-startup-databricks-raises-us10b-investors-including-gic-hits-us62b-valuation) in a US$10 billion round in 2024. Lastly, adopters integrate AI into their existing operations, such as healthcare technology player athenahealth. Yeo especially sees adopters providing the “the next big leg-up in value creation” in the coming years. “These are the large MNCs (multinational corporations), some of whom are actively experimenting, building use cases and adopting AI to increase efficiency (and) productivity,” he said, noting that AI can also change manufacturing processes. ## Picking the winners To further assess each opportunity, GIC focuses on three factors: the company’s “moat”, or structural advantages; its management; and whether it has momentum – or early wins that compound through better data and performance. As an example, Yeo outlined how Anthropic meets these three factors. First, the company has a strong moat, with the quality of its model on the back of research and development efforts, as well as its “frontier lab leadership”. On the management front, Yeo said that GIC knows the leaders of Anthropic well and believes that they can grow the business. GIC is also seeing good momentum in how Anthropic is building use cases. “It is working very closely with global companies in order to help integrate the capabilities of its model into their respective businesses,” he said. Asked whether open-source AI models from China could commoditise this space, Yeo said that having a variety of models benefits the entire ecosystem. “It’s going to increase the velocity of AI adoption, because as costs come down, we are going to see greater adoption and build of use cases. So... the market size can... grow even faster.” ## Energy and resource challenge Yeo acknowledged that a key constraint lies in AI’s resource demands, with frontier models requiring “huge amounts of compute, memory, networking and electricity”. These challenges “make the path ahead less linear and long-term outlook less clear”. However, GIC also sees new opportunities to fund critical infrastructure, particularly in renewable energy and AI-related applications that increase power grid efficiency. Data centres – in which GIC has invested since 2018 – are creating demand for onsite clean power solutions, also known as behind-the-meter solutions. These often involve solar power and battery systems, and support data centres’ power needs when grid supply is insufficient. “There are quite a number of behind-the-meter solutions that are being built, and we are... actively assessing and have invested in some of these opportunities,” said Yeo. Transmission infrastructure is another focus, with investments including European grid operator TenneT. GIC and other institutional investors last year announced that they would invest as much as 9.5 billion euros (US$10.8 billion) into[ TenneT’s German power grid](https://www.businesstimes.com.sg/international/global/norges-gic-apg-invest-9-5-billion-euros-tennets-german-grid). This is to transmit renewable energy from northern Germany to its industrial south. “AI and power supply are actually intricately linked, and AI can help to increase the efficiency of power transmission and deployment,” said Yeo. ## Art and science While closely watching AI, GIC is also looking out for “neglected” non-AI sectors where valuations could be getting cheaper. It is also watching out for uneven AI adoption across industries – which could lead to a “clear divide between AI leaders and laggards” . Another focus is on how AI could disrupt traditional sectors such as manufacturing, and erode the valuations of incumbent players. Ultimately, GIC will have to “continually re-underwrite the assets and reassess – because the technology is evolving, profit pools are shifting, and it’s still a bit unknown”, said Yeo. As he views it, investing in AI “is where art and science come together”. Decoding Asia newsletter: your guide to navigating Asia in a new global order. Sign up here to get Decoding Asia newsletter. Delivered to your inbox. Free. Copyright SPH Media. All rights reserved.