Is the AI Boom Another Dot-Com Bubble?

Artificial intelligence has been described as the biggest technological revolution since the Internet. From Wall Street to Martin Place, investors are piling into anything labelled “AI,” with share prices soaring and capital flowing at a pace rarely seen in history. Yet, for those who remember the dot-com era, there is a sense of déjà vu. Back then, investors were also convinced that a new digital economy would change everything, and in many ways it did.

However, the crash of 2000 demonstrated the fragility of speculative excess. The central question for investors today is whether the AI boom of 2025 is just another dot-com style mania, or whether we are witnessing the emergence of something far more sustainable. The answer lies in distinguishing between hype and fundamentals; leaders and laggards; and cash-generating businesses versus speculative plays.

What are the Best AI stocks to invest in right now?

Check our buy/sell tips

The AI Boom by the Numbers

The growth of the artificial intelligence sector has been nothing short of explosive. Nvidia, which produces the high-performance chips that power generative AI models, has seen its market capitalisation surge past US$4 trillion in 2025, briefly making it the world’s most valuable company.

The sheer pace of this rise is extraordinary: in early 2023, Nvidia was valued at around US$400–500 billion, before surpassing US$1 trillion by mid-2023 and now sitting at US$4 trillion. That’s a huge increase in just over two years. Some analysts even suggest there may be further upside, forecasting continued gains if AI infrastructure spending maintains its current trajectory, though the scale of those forecasts varies.

Nvidia is not alone. Palantir Technologies, once seen as a niche defence-oriented software company, is now heavily marketed as an AI stock and has enjoyed a dramatic rerating. Advanced Micro Devices (AMD) has also surged on the expectation that it will take market share in GPUs, riding the same wave of enthusiasm.

Beyond listed equities, venture capital investment has exploded. In 2025, AI startups raised billions in record-breaking rounds. The French company Mistral AI raised funding at a valuation of about €12 billion (≈US$14 billion) within two years of its founding, while US-based Anysphere, which builds AI coding tools, reported annual recurring revenue of around half a billion dollars and was valued at nearly US$10 billion.

This is reminiscent of the IPO boom of the late 1990s. Dozens of AI-focused companies are listing globally, some without profits and others with barely any revenue. The promise of future growth, rather than current fundamentals, appears to be driving AI stocks’ valuations. Such dynamics naturally raise the question: are we building towards an artificial intelligence bubble?

Lessons from the Dot-Com Bubble

To assess whether we’re in an AI bubble like dot com, we need to revisit the lessons of the late 1990s. The dot-com boom was characterised by extraordinary excitement around internet-related businesses. By March 2000, thousands of firms had raised large amounts of capital, with estimates of total fundraising in the hundreds of billions of dollars, much of it based on little more than business plans and catchy website names. The Nasdaq Composite index rose more than fourfold between 1995 and 2000, only to collapse by nearly 80 per cent over the following two years.

Companies like Pets.com, Boo.com, and Webvan became infamous examples of businesses with no viable path to profitability. At the same time, the survivors of that era, Amazon, Google, and eBay, went on to dominate the global economy. Of course, they did cop some blows, but did survive.

The key lesson for investors was simple: disruptive technology may be real, but that does not mean every stock attached to it is a winner. Maintaining valuation discipline is essential. The dot-com crash taught markets the perils of over-extrapolating early promise. The internet did change the world, but investors who bought indiscriminately in 1999 often waited decades just to break even. That experience should weigh heavily on any discussion of the tech bubble 2025 in AI.

Similarities Between the AI Boom & the Dot-Com bubble

The parallels between the current AI cycle and the dot-com era are hard to ignore. Both were driven by a narrative of unstoppable technological transformation. Both saw valuations disconnected from profits. Both were fuelled by a fear of missing out, not just among retail investors but also among institutional asset managers under pressure to show AI exposure in their portfolios.

Just as internet firms once touted “eyeballs” and “clicks” as substitutes for profits, many AI firms today justify their valuations based on future potential. Some are valued at more than 30 times revenue despite persistent losses. The race to add “AI” into a company’s branding mirrors the late 1990s, when firms added “.com” to their names to attract investors.

The IPO boom is another similarity. A wave of early-stage AI companies are listing globally, despite fragile or unproven economics. Venture funding is flowing at record levels, leading to a flood of new entrants chasing the same themes: generative AI models, agentic frameworks, autonomous coding tools, and AI for niche verticals. This competitive intensity risks saturating markets quickly, just as online retailers did in 1999.

Finally, the concentration of gains in a few mega-caps is also reminiscent of the dot-com era. Just as Cisco, Microsoft, and AOL once dominated headlines, today it is Nvidia, Microsoft, and Alphabet that capture the lion’s share of investor attention and capital.

Key Differences Today

Despite these parallels, we believe there are important differences that make the AI boom 2025 distinct from the late 1990s internet craze. The most critical is that AI already has visible, tangible adoption. Unlike dot-com firms that promised online commerce or advertising long before consumer behaviour shifted, today’s AI applications are already delivering measurable results. From productivity tools embedded in Microsoft Office to customer service bots, healthcare diagnostics, and advanced recommendation engines, AI is not just a promise; it is an operational reality.

Another major difference is the role of Big Tech. Microsoft, Google, Amazon, and Apple are not speculative startups; they are trillion-dollar companies with vast resources, cash flows, and ecosystems. Their backing of AI provides a financial safety net and infrastructure scale that was absent in 2000. In fact, Microsoft’s multi-billion-dollar partnership with OpenAI has already influenced its product strategy, driving AI-driven features and recurring revenue streams.

Crucially, some AI leaders are already profitable. Nvidia’s GPU business is generating substantial free cash flow, making it far more robust than the loss-making dot-com IPOs of 1999. The infrastructure layer of AI, chips, cloud services, and data centres, is proving to be both capital-intensive and highly profitable for incumbents.

Capital markets are also more experienced. After living through several booms and busts, from the dot-com collapse to the Global Financial Crisis, investors and regulators exercise more caution. Transparency is greater, risk management tools are sharper, and global liquidity is deeper. This doesn’t mean a crash is impossible, but it may mean the adjustment will be more selective rather than systemic.

Risks Investors Must Watch

Even with these differences, we cannot ignore the risks. The most obvious is valuation. When companies are priced for perfection, even small disappointments can trigger sharp corrections. The danger is magnified when many firms rely on similar narratives, making the entire sector vulnerable to sentiment swings.

Overinvestment is another concern. Just as billions poured into unprofitable internet ventures in the late 1990s, the massive sums now flowing into AI may produce diminishing returns. Too many startups chasing the same opportunity can erode margins and waste capital, reducing the overall return on investment.

Regulation is also looming. Governments around the world are drafting frameworks on AI governance, data privacy, and intellectual property. These could impose costs that smaller firms cannot bear, potentially accelerating consolidation but also creating uncertainty for investors.

The macroeconomic backdrop is another wildcard. The current AI enthusiasm has been supported by expectations of lower interest rates and abundant liquidity. Should inflation persist or rates remain high, the willingness of investors to pay extreme multiples could vanish quickly. Geopolitical shocks, from US-China tensions over semiconductors to conflicts disrupting supply chains, could also destabilise the sector.

Finally, technological disruption itself is a risk. AI is evolving at a breakneck pace, and today’s leading model could be obsolete in two years. Investors who back the wrong standard may find themselves holding equity in firms with no competitive edge.

A Framework for Investors

How then should investors approach the AI hype vs fundamentals dilemma? We believe the best strategy is segmentation. At the infrastructure level, chips, data centres, and cloud platforms, the moat is strong, and firms like Nvidia and Microsoft are well-positioned. These are less likely to collapse even if valuations compress. At the platform level, providers of large language models and APIs have an opportunity, but also a significant competitive risk.

At the application level, startups building tools for specific industries, the risks are highest, as differentiation is limited and capital is plentiful. Investors should anchor decisions to fundamentals: revenue growth, margins, customer stickiness, and cash generation. Lofty “total addressable market” claims are not enough. History shows that when bubbles deflate, only firms with genuine business models survive.

Position sizing and diversification are also critical. Exposure to AI should be balanced with defensive holdings, rather than being a portfolio’s sole driver. Scaling positions as companies hit milestones is one way to manage risk while staying exposed to upside. Above all, investors must remain agile. When sentiment shifts in a tech bubble 2025, the correction can be swift.

Conclusion: Bubble or Building Phase?

So, is the AI boom another dot-com bubble? The answer is nuanced. Elements of an artificial intelligence bubble certainly exist: extreme valuations, speculative IPOs, and capital inflows driven by hype rather than fundamentals. But unlike 1999, AI already demonstrates real-world adoption and is backed by financially strong incumbents.

The next two to three years will separate the winners from the pretenders. Some companies will collapse as expectations prove unrealistic, but others will entrench themselves as foundational pillars of the digital economy. For investors, the challenge is to distinguish hype from reality, and to accept that while bubbles inflate around transformational technologies, they also create opportunities for disciplined capital.

FAQs

• Is the AI boom 2025 the same as the dot-com bubble?

  Not exactly. While there are signs of speculation and overvaluation similar to the dot-com era, AI has real-world adoption and support from profitable giants like Microsoft and Nvidia. The risk is more of a selective correction than a total collapse.

• Which AI stocks’ valuations look stretched right now?

  Many smaller or mid-cap AI firms trade at price-to-sales multiples above 30× despite lacking profits. That level of valuation stretch suggests investors should be cautious and focus on fundamentals such as revenue growth and cash flow.

• Could an artificial intelligence bubble hurt big tech companies?

  The impact on large firms such as Nvidia, Microsoft, and Alphabet is likely to be limited. Their strong cash flows, entrenched ecosystems, and leadership in AI infrastructure make them more resilient, even if speculative players struggle.

• What risks could trigger a tech bubble 2025 correction?

  Higher interest rates, regulatory crackdowns on AI governance or data use, and slower-than-expected adoption could pressure valuations. Geopolitical tensions and supply chain disruptions in semiconductors also remain significant risks for the sector.

• How can investors balance AI hype vs fundamentals?

  The key is discipline. Investors should assess revenue growth, margins, customer stickiness, and cash generation instead of relying on projections alone. Diversifying exposure and scaling positions as companies hit milestones can help manage risk in this fast-moving sector.