The $800bn Gap: Do Data Centres and AI Need More Investment?
Bain & Company’s (Bain's) sixth annual Global Technology Report highlights the scale of investment required to meet AI compute demand by 2030.
The consultancy estimates that US$2 trillion in annual revenue will be needed to fund the expansion of global data centre capacity, with demand expected to reach 200GW. Even with AI-related savings factored in, Bain calculates that a US$800bn gap remains.
AI demand outpaces infrastructure growth
David Crawford, Chairman of Bain’s Global Technology Practice, describes the scale of the challenge.
“If the current scaling laws hold, AI will increasingly strain supply chains globally,” he says. “By 2030, technology executives will be faced with the challenge of deploying about US$500bn in capital expenditures and finding about US$2tn in new revenue to profitably meet demand.
“Meanwhile, because AI compute demand is outpacing semiconductor efficiency, the trends call for dramatic increases in power supply on grids that have not added capacity for decades.
“Add the arms race dynamic between nations and leading providers, and the potential for overbuild and underbuild has never been more challenging to navigate. Working through the potential for innovation, infrastructure, supply shortages and algorithmic gains is critical to navigate the next few years.”
Bain’s report underlines that AI compute demand is growing at more than twice the rate of Moore’s Law. The US is expected to account for half of the 200GW required, straining not only capital budgets but also electricity grids.
From experimentation to scaling
While some organisations are already realising EBITDA gains of 10–25% from AI deployments, most companies remain in experimentation mode.
Bain notes that leading enterprises are pushing into agentic AI, developing platforms that enable autonomous workflows across multiple systems. These require data centres capable of supporting high levels of virtualisation, low-latency interconnections and seamless access to real-time data.
The consultancy identifies four stages of agentic AI maturity, from single-task workflows to multi-agent constellations. The middle levels, where capital investment and innovation are converging, will be especially demanding for data centre infrastructure.
SaaS and sovereign AI pressures
The report also points to disruption across the SaaS sector. While AI could expand total addressable markets for SaaS providers, it will require strategic shifts in data ownership, monetisation and integration. Data centres will play a central role in enabling SaaS firms to embed AI into workflows at scale.
Anne Hoecker, Head of Bain’s Global Technology Practice, emphasises the geopolitical pressures now shaping digital infrastructure.
“Sovereign AI capabilities are increasingly seen as a strategic advantage on par with economic and military strength,” she says. “While sovereign AI is a global priority, individual countries' goals vary. Therefore, for most countries, achieving full-stack independence is not feasible, at least not today. Considering these differences, global AI standards are unlikely to converge.
“To succeed, multinational firms will need to localise not just compliance, but also their technology architecture. Businesses need to make decisions with optionality, moving boldly where confidence is high and prioritising flexibility where uncertainty rules.”
Fragmenting semiconductor supply chains add further complexity, with the US and China driving decoupling efforts and other nations struggling to balance competitiveness with sovereignty.
Beyond AI: quantum and robotics
Bain also explores adjacent technologies. Quantum computing could create up to US$250bn in value across industries including pharmaceuticals, logistics and finance, though fully fault-tolerant machines remain years away.
Humanoid robotics is also attracting significant investment, but Bain cautions that most deployments are still early stage and heavily dependent on human supervision.
Implications for investors
Despite headwinds, Bain notes that private equity investment in technology remains resilient, even as deal activity slowed in the second half of 2025.
Investors continue to view data centres and AI infrastructure as critical growth markets, though the rising capital intensity of projects will require careful allocation and strategic partnerships.
The report makes clear that meeting AI’s rising compute needs will demand unprecedented levels of collaboration between technology providers, governments, investors and utilities. Without an injection of new revenue and capital, the data centres of 2030 may fall short of powering the AI-driven economy that enterprises are racing to build.
