23 September 2025

The report projects that by 2030, global demand for AI computing could require as much as US$500 billion annually in data centre investments, translating into approximately US$2 trillion in new cloud revenue. Despite expected cost savings from AI adoption, Bain estimates that the industry will still face an annual funding shortfall of around US$800 billion to meet future demand.

The research indicates that the United States will dominate the AI computing capacity, with half of the anticipated 200 gigawatt global demand expected to be concentrated there by 2030. Even if U.S. firms reallocate all on-premise IT spending to cloud infrastructure and reinvest AI-driven efficiencies into other business areas, the funding gap would remain significant. Bain warns that AI compute demand is growing at over twice the rate predicted by Moore’s Law, putting immense pressure on energy infrastructure and supply chains, and raising concerns about potential over- or underbuilding amid geopolitical tensions.

David Crawford, Bain’s Chairman of the Global Technology Practice, emphasized that AI’s rapid scaling will challenge supply chains worldwide, requiring about US$500 billion in capital expenditures and generating US$2 trillion in additional revenue to satisfy demand. He highlighted that AI’s outpacing of semiconductor efficiency will necessitate substantial increases in power supply, often on grids that haven't expanded in decades. The report underscores the complexity of balancing innovation with infrastructure, supply shortages, and algorithmic advancements, especially amid international competition.

The report also notes that companies advancing AI across core workflows have seen profit margins increase between 10% and 25%, though many organizations remain in experimental phases with only modest productivity gains. Bain highlights the rapid development of agentic AI — systems capable of performing single tasks or orchestrating across multiple systems — with over 5% to 10% of future technology budgets potentially shifting toward foundational AI capabilities like agent platforms and security frameworks within the next few years.

Agentic AI maturity is categorized into four stages, from large language model (LLM)-powered information retrieval to multi-agent systems. Currently, most investments focus on workflow automation and orchestration systems, but Bain warns that uneven progress among industry players could widen the gap between leaders and laggards.

In the SaaS sector, Bain observes mounting challenges from generative and agentic AI but sees opportunities for market expansion. SaaS providers are encouraged to explore AI-driven automation and operational integration, with advantages in data control, setting industry standards, and adopting outcome-based pricing models to stay competitive.

The report highlights increasing fragmentation in global supply chains, driven by tariffs, export controls, and national security concerns, especially regarding sovereign AI capabilities. Bain notes that the U.S. and China are central to this trend, with China currently producing approximately 20% of the world's chips. Anne Hoecker, head of Bain’s Global Technology Practice, stressed that sovereignty in AI is viewed as a strategic asset comparable to military and economic power, which makes achieving full independence challenging. She advises multinational companies to localize compliance and technology architectures, balancing confidence and flexibility amid geopolitical uncertainties.

Bain also discusses the gradual commercialization of quantum computing, estimating a market value of up to US$250 billion across sectors like pharmaceuticals, finance, logistics, and materials science. However, widespread practical use depends on developing fault-tolerant quantum computers — a goal still several years away.

Humanoid robotics is another area of focus, with startups attracting US$2.5 billion in venture funding. Bain predicts initial deployment at scale in warehouses and logistics, with expansion into service roles such as elder care over the next decade. Most current deployments are early-stage, requiring significant human oversight.

On the investment front, technology remains resilient, accounting for 22% of North American buyouts in the first half of 2025 and supported by US$476 billion in available capital. Nevertheless, deal momentum slowed in the latter half of the year due to tariffs and geopolitical tensions. Bain notes that investors are increasingly focusing on AI-driven transformations, outcome-based contracts, and operational efficiencies as growth in traditional software sectors begins to slow.