Cathie Wood
ARK Invest chief executive Cathie Wood says an AI slowdown could ultimately make the industry stronger, despite a bruising Wall Street sell-off Caroline Wood/Wikimedia Commons CC BY-SA 4.0

ARK Invest chief executive Cathie Wood believes a slowdown in artificial intelligence infrastructure spending could leave the industry stronger, even after a 5 November sell-off erased about $3 trillion (€2.68 trillion and £2.24 trillion) from the combined market value of S&P 500 companies.

Speaking in her November In the Know commentary, published on 7 November, Wood argued that shortages of electricity, chips and other resources could force technology companies to spend more carefully. Rather than ending the AI boom, she believes the pressure could extend it.

Her comments followed a turbulent week on Wall Street. Democrats won control of both chambers of Congress in the 3 November midterm elections, prompting concern that greater scrutiny of data centres could slow construction. The incoming Democratic majorities are not due to take office until January.

On 5 November, the S&P 500 fell 4.36% while the Nasdaq Composite dropped 5.52%, its sharpest daily decline since April 2025. Nvidia shares fell more than 9%, and other companies supplying AI infrastructure were hit particularly hard.

The losses reflected fears about future spending rather than the implementation of an immediate nationwide construction ban.

Why Wood Believes a Slowdown Could Help AI

Wood's case rests on a distinction between investing quickly and investing well. Companies have committed heavily to the computing facilities needed to train and operate increasingly powerful AI systems. Bottlenecks in power supplies, semiconductors and construction could make that expansion more difficult.

But Wood argues constraints may also discourage waste. Introducing efficiencies early, she said, could make the technology revolution 'longer-lasting and much healthier'. She contrasted that possibility with the excessive investment and subsequent collapse of parts of the technology industry during the internet bubble.

That is a forecast, not a guarantee. A delayed data centre could mean postponed orders for equipment suppliers and a longer wait for returns on capital already committed. Investors are also weighing whether future regulation will restrict expansion. No uniform federal halt to AI construction has been established.

Wood pointed to credit default swap spreads as one sign of changing investor expectations. Those contracts are used to insure against default, and their prices can reflect perceived credit risk.

She said spreads linked to major technology companies had improved slightly following the election. Her interpretation was that investors anticipated a more measured pace of spending by companies building large computing networks. Narrower spreads, however, do not prove their financial risks have disappeared.

Wood's Optimism Extends Beyond Technology

Wood expects productivity gains from AI to benefit workers and businesses alike. In her view, companies could use savings to raise pay, strengthen profit margins, invest more and potentially reduce prices. How those gains are shared remains uncertain.

She also argued that the US economy was 'on a growth trajectory that is much stronger than most people recognize'. The government's advance estimate, released on 29 October, showed real gross domestic product growing at an annualised 3.4% rate in the third quarter.

Housing remains weak, but Wood anticipates an eventual recovery. She also expects cheaper AI, improved productivity and lower oil prices to relieve inflationary pressure, describing the possible combination as a 'win-win situation'.

For now, markets are focused on what slower construction might cost. Wood is making a different bet: that a pause in the race to build could prevent a more damaging reckoning later.