US stock markets suffered their sharpest weekly decline in months on Friday, with the Nasdaq falling over four percent in its biggest single-day drop since April 2025. According to BBC Business reporting, the selloff was triggered by a stronger-than-expected jobs report that spooked investors with the prospect of continued Federal Reserve tightening — but the mechanism of the fall was concentrated in technology and AI-adjacent equities, with major investment funds rotating out of chip and AI companies into healthcare, utilities, and consumer staples. The S&P 500 fell 2.6 percent and the Dow Jones lost 1.35 percent. Bitcoin also dropped sharply. President Trump, who had been expecting markets to celebrate strong employment data, expressed frustration: “I hope the market starts to learn that when you have good numbers the market should go up not down.” Meanwhile, Trump separately invited top AI executives to the White House to discuss a proposal for the US government to take public stakes in their companies.
The received wisdom
The conventional reading of Friday’s selloff is that this is a classic interest-rate story: strong employment data reduces the probability of Federal Reserve rate cuts, which increases the discount rate applied to future earnings, which hammers long-duration growth stocks — and AI companies, which are expected to generate most of their profits far in the future, are exactly the kind of stocks that suffer most from this repricing. In this reading, the fundamentals of AI investment remain sound. The companies building large language models and semiconductor infrastructure are genuine engines of future productivity growth. The selloff is a rational recalibration of timing, not a verdict on the underlying technology. Defenders of current AI valuations point to real and accelerating adoption in enterprise software, drug discovery, logistics optimisation, and content generation. The argument is that previous technology bubbles — the dotcom era chief among them — did eventually produce the productivity gains investors anticipated; it just took longer and required a shakeout that eliminated weaker players first.
A different read
The dotcom comparison, deployed by bulls as consolation, should give pause rather than comfort. The dotcom bubble involved genuine underlying technology that transformed the economy — and yet investors who bought at peak 2000 valuations spent over a decade waiting to break even on the Nasdaq. The question is not whether AI is real; it demonstrably is. The question is what a defensible earnings multiple looks like for companies that are burning capital at extraordinary rates while the competitive landscape remains radically uncertain.
BBC Business has been tracking the AI bubble question closely, noting the degree to which a handful of mega-cap technology firms now account for a disproportionate share of the entire US equity market’s value. When those firms move, the indices move. This concentration risk is not new — it has been a structural feature of US markets since 2020 — but it creates fragility. A handful of negative data points about AI adoption rates, regulatory interference, or competition from Chinese models could trigger corrections far more severe than Friday’s.
There is a second-order political story here that deserves attention. Trump’s proposal to have the US government take public stakes in AI companies, announced on the same day markets sold off, sits oddly with free-market orthodoxy. The Anthropic co-founder’s call this week for AI labs to pause development and accept greater human oversight — and the firm’s warning to Al Jazeera that humans risk losing control of the technology — frames the policy environment in which AI companies are operating. Government equity stakes in technology companies would represent a significant departure from the American model of arms-length market development and a shift toward the kind of industrial policy the right has traditionally resisted.
The deeper structural question is whether the current wave of AI investment is building productivity infrastructure that will generate broad economic returns, or whether it is primarily enriching a narrow group of capital holders and equity-compensation recipients in ways that do not diffuse through the wider economy. The jobs report that spooked markets on Friday showed strong employment — 172,000 jobs added, heavily concentrated in restaurants, bars, and hotels as the US prepares for the World Cup. That employment picture — service sector strength, tech sector uncertainty — is the bifurcated economy that AI investment, so far, has not resolved. The promise of AI is eventually to raise productivity broadly. The current evidence is that it has raised asset prices broadly and productivity narrowly.
A right-leaning reading would add: the Federal Reserve’s continued high-rate posture, which is what triggered Friday’s rout, is itself a consequence of inflationary fiscal policy. The combination of large deficit spending and unprecedented AI investment capital is running into a monetary tightening regime that is cooling the enthusiasm for leveraged bets on long-horizon technology payoffs. The appropriate response is not government equity stakes in AI companies or monetary loosening to protect equity values. It is fiscal discipline that reduces the inflationary pressure requiring high rates in the first place.
What to watch
Watch the Federal Reserve’s next meeting statement for signals about the rate path — any indication that the jobs strength is being read as inflationary will accelerate the rotation out of growth equities. Watch Nvidia’s next earnings report, which functions as a proxy for real enterprise AI spending rather than announced intentions. If capital expenditure at hyperscalers continues to grow, the investment thesis remains credible; if it plateaus, the correction could deepen. And watch the outcome of Trump’s White House AI summit: if the government equity stake proposal advances, it will provoke a genuine policy debate about whether America’s technology advantage is best secured through market competition or through a new form of industrial policy.
— J