Technology Sector Warns of AI Bubble with Risks Like 2008

The tech sector is seeing mounting warnings of overvaluation on a scale reminiscent of the dot-com bubble or even the 2008 financial crisis. The Bank of England already warned in December 2025 of a possible bursting of the AI bubble, noting that British share prices are near the "most extreme" levels since the global financial crisis of 2008. In the US, valuations are reaching dimensions comparable to the period before the dot-com bubble burst.

The Scale of the Current Investment Wave

Major technology companies have invested several trillion dollars in artificial intelligence and related infrastructure. This sum is referred to in analyses as a 3-trillion-dollar bet that could trigger a crash on the order of 2008 if it fails. Investments are flowing primarily into data centers, AI models and associated hardware infrastructure.

Valuation metrics in the technology sector are now at levels comparable to the dot-com era. The price-to-earnings growth ratio (PEG ratio) suggests the rally has limited room to move upward. Analysts who had predicted the bursting of the dot-com bubble are now warning again of a correction that could be worse than the 2008 financial crisis. One such analyst had previously predicted that the S&P 500 could fall by up to 75 percent from its peaks – a steeper decline than during the financial crisis.

Two Possible Scenarios for a Correction

Analysts distinguish between two fundamentally different scenarios for how the bursting of the AI bubble could unfold. The first scenario follows the pattern of the dot-com bubble of the late 1990s. Back then, overinvestment led to the collapse of hundreds of companies and a brief recession. In the long term, however, the period had the advantage that massive investments in internet infrastructure advanced the technology sector overall. The dot-com crisis caused a mild recession, whose effects were far less devastating than those of the subsequent 2008 recession, which was triggered by the collapse of the market for mortgage-backed securities and the housing bubble.

The second scenario follows the pattern of the 2008 financial crisis. In this case, the bursting of the bubble would not only hit the technology sector, but would drag the global economy down with it. The systemic interconnection of major technology companies with other sectors, as well as their importance for stock indices and institutional portfolios, could trigger a domino effect.

Gradual Corrections Rather Than Sudden Crash

In addition to these two extreme scenarios, a third variant is also considered likely: a series of smaller corrections instead of a single catastrophic collapse. Investors and funds could successively overextend their positions in various tech trends, thereby creating multiple successive bubbles in subsectors. Historical examples show that such patterns have already occurred in the past – for instance, with short-lived hype around mobility companies such as e-scooter providers.

Historical Context and Regulatory Attention

In March 2025, the bursting of the technology bubble marked its 25th anniversary. Goldman Sachs used this milestone as an opportunity to analyse those historical developments and draw parallels to the present. The current situation shows both similarities and differences to historical bubbles. While the dot-com bubble was characterised by investments in internet infrastructure, today's investments are focused on artificial intelligence and machine learning.

Regulatory authorities such as the Bank of England are actively monitoring valuation metrics and comparing them with benchmarks from past crises. The concern focuses on valuations not being backed by corresponding earnings or real business models. Unlike the dot-com bubble, when many start-ups without viable business models rushed to go public, today's main players are established large corporations. Their market capitalisation and weight in indices make them systemically relevant.

Risks and Uncertainties

The central question remains whether massive investments in artificial intelligence will generate sufficient returns in the foreseeable future to justify valuations. Should the hoped-for breakthroughs fail to materialise or be delayed, a revaluation of shares threatens. The warning of a "doom loop" – a self-reinforcing downward spiral – is based on the fear that falling share prices could lead to loss of confidence, capital outflows and further price declines.

Investors face the challenge of distinguishing between justified growth opportunities and speculative excess. Historical comparisons show that both mild and severe corrections are possible. The coming quarters will show whether current valuations can be supported by real business results or whether a correction is inevitable.