Artificial intelligence has become the biggest investment story of the decade.
Technology companies are pouring hundreds of billions of dollars into AI chips, cloud infrastructure and data centers, while investors continue to push the sector to record valuations. Every major breakthrough appears to strengthen the belief that AI will define the next era of economic growth.
But beneath the optimism, some of the world’s most influential financial institutions are beginning to sound a note of caution.
Economists at the Bank for International Settlements (BIS), often described as the “central bank for central banks,” believe parts of today’s AI investment boom are beginning to resemble the early stages of previous financial bubbles.
The warning is not that artificial intelligence lacks value. It is that investor expectations may be rising faster than the industry’s ability to generate sustainable returns.
That distinction could prove critical in the years ahead.
The World’s Biggest AI Investment Wave
The scale of today’s AI spending is unlike anything the technology sector has experienced in recent history.
The world’s largest cloud providers and technology companies are expected to spend close to $1 trillion on AI infrastructure by the end of the year. The money is funding advanced graphics processors, new data centers, networking equipment and the enormous computing power required to train increasingly capable AI models.
Supporters argue that these investments are laying the foundation for the next industrial revolution.
History certainly offers reasons for optimism. Railways, electricity and the internet all experienced periods of heavy investment before transforming economies around the world.
Yet history also offers another lesson.
Transformational technologies often attract speculative investment long before profitable business models fully emerge.
That is the pattern central bankers are watching closely.
The Revenue Cycle Behind the AI Boom
One of the BIS’s biggest concerns is how money is moving within the AI ecosystem.
Many of the largest technology companies are investing billions of dollars in AI startups. Those startups frequently spend much of that funding purchasing cloud computing services from the very companies that invested in them.
The transactions are genuine.
The revenue is real.
However, economists worry that this creates a self-reinforcing investment cycle.
The cloud provider reports rising revenue. The startup demonstrates rapid growth. Investors reward both companies with higher valuations, encouraging even more investment across the sector.
The concern is not that companies are inflating their numbers.
It is whether a meaningful share of today’s revenue depends on continued investor funding rather than long-term customer demand.
If investment slows before commercial adoption reaches the same pace, growth could weaken much faster than markets currently expect.
Why Shadow Banking Matters
Building AI infrastructure requires extraordinary amounts of capital.
Not all of that financing comes from traditional banks.
A growing share is being provided by private credit firms, investment funds and other non-bank financial institutions that operate with fewer regulatory requirements. Together, these institutions form what economists refer to as the shadow banking system.
Shadow banking plays an important role in financing economic growth.
However, it is generally less transparent than the traditional banking sector, making it harder for regulators to measure how risk is spreading across the financial system.
According to the BIS, exposure to AI and information technology assets has increased significantly among these lenders.
If confidence in the sector were to weaken, those interconnected financial relationships could amplify market stress far beyond the technology industry itself.
Why This Boom Is Different From Previous Ones
Every generation believes its defining technology is different.
Sometimes it is.
Sometimes the financial behavior surrounding it is surprisingly familiar.
When central bankers compared today’s AI investment surge with some of history’s most famous speculative booms, one finding stood out.
The speed of investment growth has been extraordinary.
| Technology Boom | Investment Growth | Time Period |
|---|---|---|
| Current AI Boom | 4.5× | 3 years |
| Canal Mania | 4.1× | 5 years |
| British Railway Mania | 2.7× | 4 years |
| Dot-com Bubble | 1.9× | 5 years |
Rapid investment does not automatically signal a bubble.
However, financial history suggests that when capital flows into an industry faster than profitable demand develops, markets become increasingly vulnerable to sharp corrections.
That is precisely why central bankers are paying attention.
The Physical Limits of Artificial Intelligence
Unlike many previous software revolutions, artificial intelligence depends heavily on physical infrastructure.
Training advanced AI models requires specialized chips, enormous amounts of electricity, sophisticated cooling systems and a constant expansion of data center capacity.
Those resources are neither cheap nor unlimited.
Electricity demand is already rising rapidly in regions hosting major AI facilities, while advanced chip manufacturing remains concentrated among only a handful of suppliers.
There is another challenge.
High-performance AI hardware becomes obsolete within a few years, forcing companies to replace expensive equipment on a continual basis.
This creates a business environment where infrastructure costs keep rising, even as companies are still trying to prove that customers will consistently pay enough to justify those investments.
Does This Mean an AI Crash Is Inevitable?
No.
That is not what the BIS is saying.
Central bankers are not predicting another global financial crisis. They are warning against assuming that every dollar invested in artificial intelligence will automatically produce profitable returns.
History shows that speculative bubbles and transformative technologies often exist at the same time.
The railway boom ended in financial losses for many investors, yet the railway network transformed transportation.
The dot-com bubble erased billions of dollars in market value, but the internet still reshaped the global economy.
Artificial intelligence could follow a similar path.
Some companies may fail.
Others may become the defining businesses of the next generation.
The difficult task for investors is separating genuine long-term value from short-term market enthusiasm.
The Bigger Picture
Artificial intelligence is already improving scientific research, business productivity and software development.
Few economists question its long-term importance.
The debate is about something else.
Can financial markets remain patient while the technology matures, or have expectations already moved ahead of economic reality?
That is the question central bankers want investors to consider.
If AI adoption continues to expand and companies build sustainable businesses, today’s investment boom may ultimately prove justified.
If spending continues to outpace commercial returns, however, the industry could experience a painful correction before reaching a more stable foundation.
History rarely punishes innovation.
It does, however, have a long record of correcting excessive optimism.
That may be the most important lesson behind the BIS warning.
Artificial intelligence has the potential to reshape the global economy.
Whether today’s extraordinary investment boom becomes a historic success or another chapter in financial history will depend not only on technological progress, but also on whether expectations remain grounded in economic reality.
READ ALSO: Instagram Shifts Control to Users with Test of Brand-New ‘Your Algorithm’ Feature
