USA AI Euphoria Echoes The Internet Bubble

Despite a growing list of economic risks, from tariffs and immigration curbs to soaring debt and stubborn inflation, corporate America and its investors are betting heavily on one idea: that artificial intelligence can outgrow every problem.

Over the past year, that optimism has become a self-fulfilling force. US companies are spending hundreds of billions on AI infrastructure, tools and chips, driving roughly 40 per cent of all GDP growth in 2025. Some economists believe the true contribution could be even higher once indirect effects are counted.

The stock market tells a similar story. AI-linked firms have delivered about 80 per cent of the total gains in US equities this year, drawing capital from every continent and fuelling a consumption boom among the wealthy. The richest 10 per cent of Americans, who own 85 per cent of US stocks, now generate half of all consumer spending, the largest share on record. The nation’s economy, increasingly, rides on the confidence and spending power of those who have profited most from the AI rally.

Without that momentum, the outlook would be less convincing. Beneath the surface, the same vulnerabilities that haunted earlier cycles remain: a shrinking workforce, fiscal excess, and geopolitical tension.

Immigration and Labour Pressures

No other developed nation has experienced an immigration whiplash on the scale of America’s. Net arrivals surged to more than 3 million in 2023, only to collapse this year to an estimated 400,000 after the Trump administration tightened entry. Goldman Sachs calculates that this alone could shave more than 20 per cent off America’s long-term growth potential.

Yet Wall Street appears unbothered. The prevailing view is that AI will offset labour shortages by raising productivity and automating tasks once done by people. As one investor put it recently, “labour scarcity is now a design opportunity.”

Debt, Deficits and Faith in Growth

Public finances tell a starker story. US government debt sits near 100 per cent of GDP, levels last seen after the Second World War, and continues to rise. In any other country, such numbers might alarm bond markets. Instead, investors seem to believe that if AI delivers even a modest productivity miracle, the burden will stabilise as national income rises.

Bond traders have acted accordingly. Yields on US 10-year Treasuries have fallen this year, even as investors sold off the debt of Japan, France and the UK, countries running smaller deficits. In global markets, faith in American innovation continues to outweigh concern over American arithmetic.

Productivity Dreams

The belief that AI will transform output per worker sits at the heart of this enthusiasm. Greater efficiency promises to lift GDP, ease inflationary pressure, and allow wages to rise without stoking prices. Policymakers and investors alike are eager to imagine that outcome.

To date, US productivity has indeed outpaced most of the developed world. That edge, combined with vast corporate investment in AI infrastructure, has reinforced the conviction that the United States is building a lead others cannot match.

Foreign investors have followed the story with their wallets. In the second quarter alone, non-US buyers poured a record $290 billion into American equities. Overseas ownership now accounts for nearly 30 per cent of the US stock market, the highest share since 1945. Even as Europeans shun US goods over tariff disputes, they continue to buy US technology stocks in bulk.

The Dollar Exception

The only weak note in the “buy America” refrain has been the dollar, which has softened in recent months. Analysts largely attribute that to foreign investors hedging currency exposure rather than withdrawing confidence.

One Big Bet

Taken together, these trends make the United States today resemble one enormous wager on AI, its promise of productivity, profit and growth. Strip out the tech sector, and global peers have begun to outperform. Every major industry from utilities and healthcare to banking has done better in Europe or Asia this year than in the US. That divergence underscores how narrow the foundations of the American boom have become.

If the AI story falters, whether through slower adoption, regulatory backlash or technological disappointment, the consequences could be severe. A stock-market correction would drain the wealth effect that now sustains consumer spending. A fall in confidence could expose fiscal weaknesses long ignored.

Lessons From 1999

The last time investors believed technology alone could defy gravity was in the late 1990s. Back then, the internet promised to reinvent business and society. Capital flooded into unproven companies, valuations soared, and analysts declared a “new economy” immune to old-fashioned cycles. When reality caught up in 2000, the Nasdaq lost nearly 80 per cent of its value, and it took more than a decade to recover.

The parallel with today’s AI boom is imperfect, as many of the leading firms now are profitable giants with genuine products and cashflow. But the psychology feels familiar. In 1999, too, productivity was said to be on the verge of a revolution. It did arrive eventually, but only years later and at far lower rates than the optimists had forecast.

A second lesson lies in concentration. During the dot-com era, market gains were heavily skewed toward a handful of names. When sentiment turned, diversification offered little shelter. The same dynamic now applies to the so-called “AI Magnificent Seven,” which dominate both indices and investor imagination.

Third, the internet bubble showed how technological promise can mask deeper economic strains. In 1999, the US ran rising deficits and relied on capital inflows from abroad. The pattern is repeating: foreign money finances American consumption, drawn by belief in innovation rather than by trade surpluses or fiscal prudence.

The Road Ahead

Few doubt that AI will transform industries and reshape productivity over time. The question is one of scale and timing. If the gains arrive more slowly than investors expect, markets may once again face a painful repricing.

For policymakers, the challenge is to ensure that the AI boom feeds into broad-based growth rather than another narrow surge in asset prices. For investors, the discipline lies in remembering that even the most revolutionary technology cannot abolish cycles of exuberance and correction.

In 1999, euphoria met reality in the space of a year. Whether 2025 proves different will depend less on the brilliance of machines than on the restraint of the humans investing in them.