The froth is rising in Silicon Valley, and the City of London is watching with alarm. A growing chorus of analysts now warns that the artificial intelligence stock rally has entered dangerous territory, echoing the dot-com mania that preceded the 2000 crash. This week, the tech-heavy Nasdaq Composite suffered its worst single-day drop in months after a routine earnings miss from a major cloud provider – a tremor that many believe is the first crack in a dangerously overvalued sector.
‘We are witnessing a classic case of irrational exuberance,’ said Helena Cross, head of technology equity research at Barclays. ‘Companies are being valued on promises of future AI dominance rather than actual revenue streams. The multiples are unsustainable.’ Indeed, the average price-to-earnings ratio for AI-focused firms now sits above 60, compared to 20 for the broader S&P 500. Startups with no clear path to profitability are trading at valuations that imply they will capture entire industries overnight.
The root of the concern lies in the gap between hype and reality. Generative AI tools, while impressive, have not yet translated into the productivity gains that would justify the trillions of dollars in market capitalisation. ‘We are in the trough of disillusionment,’ said Dr. Anika Patel, a digital ethics fellow at the London School of Economics. ‘The technology is real, but the business models are not. Investors are betting on a future that may be five to ten years away, ignoring the current costs of compute, energy, and data.’
The parallels with the dot-com era are striking. Then, companies with ‘.com’ in their names saw their stocks soar before crashing spectacularly. Today, adding ‘AI’ to a pitch deck can double a valuation overnight. The pattern is eerily similar: a transformative technology, low interest rates, and a herd mentality among venture capitalists and retail traders alike. ‘The difference is that AI is truly disruptive,’ noted Julian Vane, Technology & Innovation Lead at a London-based think tank. ‘But that doesn’t mean the market pricing is correct. In the 1990s, the internet changed everything – but most of the companies that skyrocketed went bankrupt. The survivors are the ones we use today. The same will happen here.’
Regulators are beginning to stir. In the US, the Securities and Exchange Commission has signalled it will scrutinise AI-related claims made by listed firms. In Europe, the AI Act is forcing companies to be more transparent about their models. Yet the bubble may burst before any legislative impact is felt. Central banks, still fighting inflation, are keeping interest rates higher than many had hoped, squeezing the easy money that fuelled the rally.
So what happens next? ‘A correction is inevitable,’ said Cross. ‘The question is whether it will be a gradual deflation or a sudden crash. Given the leverage in the system and the retail frenzy, I fear the latter.’ For the average investor, this means buckle up. The AI revolution is not a mirage, but the route to it will be rocky. Those who bought at the peak may be left holding the bag. As Vane put it: ‘The future of AI is bright, but the stock market is a discounting mechanism that has priced in a utopia that cannot arrive fast enough.’
