The explosive rise of artificial intelligence (AI) continues to dominate headlines and fuel staggering levels of investment. AI companies are now valued in the hundreds of billions, with major tech firms pouring record amounts into data centers and infrastructure. As the world marvels at rapid advancements and new billionaires mint seemingly overnight, a pressing question emerges: Are we living through an AI bubble, and if so, when will it pop?

Market historians and economists are quick to draw parallels with the dot-com bubble of the late 1990s and early 2000s, when frenzied investment and euphoric speculation gave way to a dramatic crash. AI, they warn, could be on a similar trajectory. To understand what’s at stake, it’s crucial to examine what triggered the dot-com collapse, the similarities and differences in today’s AI market, and what may lie ahead for technology investors.

The Dot-Com Bubble and Its Bitter Aftermath

Bursting in March 2000, the dot-com bubble wiped out trillions of dollars in market value. The crash was precipitated by a confluence of factors, notably repeated Federal Reserve interest rate hikes in 1999 and 2000 that made speculative bets less attractive. At the same time, a recession in Japan and wider global uncertainty shook investor confidence, triggering a mass re-evaluation of tech stock valuations.

Yet, the most critical flaw was deeper: the fragile business models underpinning many internet startups. Some companies, like Commerce One and Pets.com, achieved valuations in the billions despite earning little or no revenue. Marketplace darlings such as TheGlobe.com saw stock soar then crash, supported more by buzz than by sustainable profit or cash flow.

Compounding the speculative fever was runaway infrastructure spending. Telecom companies, convinced that internet usage would skyrocket endlessly, laid over 80 million miles of fiber optic cable—much of which remained “dark,” or unused, for years after the crash. When demand fell short, shares in suppliers like Corning and Ciena collapsed by over 90%, while numerous service providers went bankrupt or merged at bargain prices.

How the AI Gold Rush Compares

Today, the AI sector shows striking similarities. Global corporate AI investment soared to $252.3 billion in 2024, marking a thirteenfold growth since 2014 according to Stanford University. Technology giants—Amazon, Google, Meta, and Microsoft—have collectively pledged to invest a record $320 billion in infrastructure this year, much of it earmarked for AI data centers and computational power.

Startups and unicorns are thriving as well. In 2025, AI created dozens of new billionaires, with the number of AI unicorns—private, venture-backed firms worth over $1 billion—nearing 500 and a combined estimated value hitting $2.7 trillion. OpenAI, propelled by the unprecedented global usage of ChatGPT and its enterprise AI offerings, is now valued at around $500 billion and could reach $20 billion in annualized revenue by 2025’s end.

This exuberance has also led to historic bets on AI infrastructure. Meta CEO Mark Zuckerberg has unveiled plans for an AI data center “so large it could cover a significant part of Manhattan.” OpenAI’s Stargate Project—backed by Oracle, SoftBank, and MGX—aims to create a $500 billion nationwide AI data center network.

Revenue Reality vs. Runaway Hype

Notably, there are important differences. Unlike their dot-com predecessors, today’s leading AI firms generally generate significant revenues. Microsoft’s Azure cloud service, with its heavy AI focus, grew 39% year-over-year to reach an $86 billion run rate. Amazon Web Services and Google Cloud are similarly pushing toward double-digit billion-dollar figures in AI-specific income streams.

Still, the gap between infrastructure spending and realized income is formidable. Over the last two years, the top five U.S. tech giants have spent an estimated $560 billion on AI capex, but have only reported about $35 billion in direct AI revenues. And while OpenAI’s and Microsoft’s growth is strong, smaller players and many startups struggle to convert large-scale pilot projects into real, sustainable customer value.

According to a recent MIT study, 95% of generative AI pilots launched by enterprises in the past year have failed to yield meaningful business impact—despite more than $40 billion invested in generative AI initiatives. This highlights the sector’s challenge: enormous capital flowing into transformative, but still immature, technologies, much like during the dot-com era.

The Skeptics and the Defenders Weigh In

Even vocal AI advocates acknowledge the risks. OpenAI CEO Sam Altman admits, “Are we in a phase where investors as a whole are overexcited about AI? My opinion is yes. Is AI the most important thing to happen in a very long time? My opinion is also yes.” This cautious optimism is echoed by leading analysts at Morgan Stanley and Goldman Sachs, who regularly report both the transformative potential and the risk of inflated expectations for AI.

Veteran investor and technology commentator Paul Kedrosky notes, “We’re seeing the biggest capital cycle in technology since the telecom buildout of the late ’90s. The lesson from that period is clear: the winners will be those who build lasting business models, not just infrastructure.”

What’s Next? Navigating the Boom—Or Bust

Whether or not the AI bubble bursts in dramatic fashion, as with dot-coms, depends on the industry’s ability to avoid past mistakes. If massive investments in data centers and AI infrastructure are supported by scalable, real-world revenue and productivity, the long-term rewards could be transformative. Conversely, if demand falls short and infrastructure goes underused, today’s “bright” fiber cables and GPU clusters could become tomorrow’s “dark assets.”

Most experts agree: AI will change the global economy. But the pace, sustainability, and value extraction from that change will determine if today’s speculative fever moderates, or if a sobering recalibration—as in 2000—awaits. Investors, founders, and policymakers would do well to remember the past, even as they build the future.