AI Infrastructure Spending Races Toward $1.6 Trillion Despite Profitability Questions

The artificial intelligence industry is experiencing unprecedented capital deployment, with spending on infrastructure projected to surge from $765 billion in 2026 to $1.6 trillion by 2031, according to Goldman Sachs analysis first reported by The Guardian.

This massive investment wave comes as major AI companies race toward public markets. SpaceX recently announced plans for a $1.77 trillion valuation, while Anthropic filed for an initial public offering, with OpenAI expected to follow. Yet beneath the market euphoria, fundamental questions about profitability and sustainability are intensifying.

Market concentration reaches historic levels

Technology stocks now dominate the S&P 500 to an unprecedented degree. Research from Bianco Research found that 41 AI-related stocks account for nearly half the index's total market value, with the S&P 500 rising nearly 80% over five years driven primarily by seven companies: Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia, and Tesla.

Analysts at Saxo UK warn this concentration creates systemic risk. "The entire market has become one giant AI edifice," said Neil Wilson, noting that current conditions—including potential 1970s-style inflation and stretched valuations—could trigger a dotcom-style crash.

Adoption accelerates while costs spiral

Corporate AI adoption has jumped from 33% in 2023 to nearly 80% today, according to McKinsey data reported by The Guardian. OpenAI's ChatGPT now reaches 1 billion monthly active users, a record for any application. Meanwhile, Anthropic's Claude has gained significant ground, with internet traffic analysis from Kentik showing Claude's user growth significantly outpacing ChatGPT between January and April 2026.

However, the economics of AI usage are becoming problematic. Token costs—the units measuring AI inputs and outputs—are rising sharply. OpenAI charges $5 per million input tokens and $30 per million output tokens for GPT-5.5. One unnamed company reportedly spent $500 million in a single month on Claude Code licenses, according to Axios.

"The costs are getting completely out of control," said Liam Betsworth, founder of British AI startup Pendra, as reported by The Guardian. Software developers are rapidly upgrading from basic to premium subscriptions as their usage scales.

Infrastructure constraints loom

Datacentre construction represents a critical bottleneck. Bloomberg estimates 23 gigawatts of capacity was under construction globally in 2025, with property firm JLL forecasting 100 gigawatts of additional capacity needed between 2026 and 2030—equivalent to 1,200 new datacentres.

Cecilia Rikap, an associate professor at University College London, questioned whether governments can deliver the necessary grid expansion and funding, particularly given environmental concerns.

Meanwhile, AI capabilities continue advancing rapidly. Research organization METR found that AI model capabilities are doubling every four months, with Anthropic's Claude Mythos achieving 50% success rates on tasks requiring eight hours to two days of human expert time.

Economic dependency deepens

Datacentre investment now props up broader economic growth. A Harvard economist calculated that information processing equipment and software investment accounted for 92% of US GDP growth in the first half of 2025, according to The Guardian's analysis. This concentration means any slowdown in AI spending could have significant economic and political consequences.

Goldman Sachs analysts acknowledged the risk: "At the scale of capital being committed, even modest delays in execution invite real scrutiny around the demand assumptions used to underwrite these investments."

Why it matters

The AI industry faces a critical inflection point where massive capital deployment must translate into measurable productivity gains and profitability. If companies cannot demonstrate sufficient return on investment to justify rising token costs and infrastructure spending, the current valuation levels—and the broader economic growth they support—become unsustainable. The concentration of market value in AI-related stocks amplifies systemic risk across financial markets.

The analysis and data were first reported by The Guardian, drawing on research from Goldman Sachs, McKinsey, Bianco Research, METR, and other sources.