Artificial intelligence is no longer a technology-sector story.

The race to build the computing capacity behind AI is reshaping American investment, trade, credit markets, electricity demand, construction and, increasingly, the prices consumers pay for technology.

The change is being driven by an extraordinary burst of spending on data centres, advanced chips, networking equipment and power systems.

Amazon, Alphabet, Microsoft and Meta are committing hundreds of billions of dollars each year to expand AI infrastructure, while Oracle, chipmakers, cloud providers, utilities, construction companies and private-credit firms are being drawn into the same investment cycle.

The four largest US technology groups have spent more than $1.1 trillion on capital expenditure since the AI boom began in 2023. Their combined spending is expected to approach three-quarters of a trillion dollars in 2026 alone.

Amazon has raised its own annual investment plan to $220 billion, largely to meet demand for cloud computing and AI services.

That spending is now large enough to influence national economic data.

Investment in computers, software and data-centre equipment has become an important source of US growth at a time when consumer spending has been less reliable.

Estimates differ, but AI-related investment has accounted for a substantial share of recent expansion in private investment and gross domestic product.

The benefit is spreading well beyond Silicon Valley.

Data-centre construction is creating demand for land, concrete, electrical equipment, cooling systems, fibre networks, generators and skilled labour.

Utilities are planning for large new loads of electricity.

Equipment makers and engineering companies are reporting stronger orders as cloud providers race to secure capacity.

But the boom is also making the American economy more dependent on a narrow set of companies and supply chains.

Much of the hardware needed for AI is imported, particularly from East Asia, where semiconductor manufacturing and advanced electronics are concentrated.

Federal Reserve research estimates that roughly 90% of equipment used by high-technology sectors is sourced abroad.

That import dependence has consequences.

A sustained AI buildout can widen the US current-account deficit because demand for foreign chips and equipment rises faster than exports.

It can also add to inflationary pressure if the price of semiconductors, memory chips and specialised computing equipment increases.

Computer and chip prices, which fell for decades, have begun rising again in parts of the supply chain.

Consumers may feel that effect even if they never use an AI chatbot.

The same memory chips, processors and manufacturing capacity required for AI servers are also used in smartphones, laptops and other electronics.

Apple has largely avoided the most expensive data-centre spending by its rivals, but rising component costs and supply constraints can still affect the price and availability of iPhones and other consumer devices.

That does not mean every price increase is caused by AI, but the infrastructure race is adding pressure to shared technology supply chains.

The financing model is changing as well.

The biggest technology groups still generate vast cash flows, yet the scale of the buildout has encouraged greater use of corporate bonds, private credit, long-term leasing and special-purpose financing vehicles.

Some arrangements allow infrastructure investors to buy chips or build data centres, then lease the capacity to AI companies.

That expands the pool of available capital, but also shifts more risk into credit markets if demand or pricing assumptions prove too optimistic.

The stock market is another channel.

Enthusiasm for AI has added trillions of dollars to the value of technology companies and increased household wealth for investors who own those shares.

That wealth can support consumer spending.

The reverse is also true: if expectations for AI revenue fall sharply, a correction in highly valued technology shares could affect confidence, investment and spending more broadly.

The core uncertainty is whether the vast infrastructure buildout will produce productivity gains and profitable services quickly enough to justify its cost.

Cloud revenue is growing strongly for several of the largest providers, and demand for computing capacity remains high.

Yet many AI products are still searching for durable business models, while the expense of chips, power and data centres is being committed years before the final returns are known.

For now, artificial intelligence is acting as an economic stimulus as much as a software revolution.

It is lifting investment and industrial demand, while exposing the United States to higher import costs, greater financial concentration and a larger bet that the next generation of computing will pay for the infrastructure now being built.