Artificial intelligence is triggering one of the largest waves of capital investment in modern economic history. Companies worldwide are set to spend about $375 billion on AI infrastructure in 2025, with investment climbing to $500 billion in 2026, according to The New York Times.
The money is being poured into data centres, semiconductor fabrication plants and energy systems, fuelling growth across the global economy. Government data suggests software and computing hardware already account for a quarter of domestic economic growth, underscoring how central AI has become to economic momentum.
Industry experts forecast global investment in AI infrastructure could reach $7 trillion over the next decade. Data centre construction is now outpacing office building projects, reflecting the strategic priority companies and governments place on enabling AI at scale.
Jensen Huang, chief executive of Nvidia, predicts the AI chip market alone will be worth trillions within five years. He expects $3 to $4 trillion in AI infrastructure spending by 2030, supported by demand from technology giants and data centre operators. Nvidia’s outlook is reinforced by record global sales of its high-end GPUs.
Tech majors are backing those projections with landmark commitments. Google has pledged $9 billion for new cloud and AI infrastructure in Virginia and secured a $10 billion cloud services contract with Meta, its biggest ever, to support Meta’s AI ambitions. Meta, in turn, is pursuing what could become the world’s largest AI data centre in Louisiana, with an estimated $50 billion price tag. It is also planning a 2,000-MW data campus and investing up to $65 billion in AI infrastructure by the end of 2025, including the purchase of more than 1.3 million GPUs.
Brookfield Asset Management forecasts $4 trillion of the next decade’s AI investment will go into chips, $2 trillion into data centres and around $500 billion each into power and supporting technologies. AI-focused data centre capacity is expected to almost double to 15 GW by the end of 2025, then expand more than fivefold to 82 GW by 2034.
The financial intensity of model development is another driver. Training frontier systems such as OpenAI’s GPT-4 and Google’s Gemini Ultra can cost tens to hundreds of millions of dollars, making access to advanced infrastructure a prerequisite for competition.
For the UK, the boom presents both opportunity and challenge. By encouraging responsible innovation, aligning investment with clean energy and embedding strong regulatory frameworks, Britain could turn the surge into a competitive advantage and bolster its ambition to be a global AI leader.
The AI infrastructure race is more than an economic trend—it represents a foundational shift in technology and industrial strategy. Nations and firms that can combine capital with foresight stand to shape the digital economy for decades to come.
Created by Amplify: AI-augmented, human-curated content.
