AI stock sell-off sends Wall Street reeling and erases midterm rally gains

U.S. stocks suffered a sharp, broad decline as investors continued selling companies linked to artificial-intelligence data centers. The S&P 500 fell 3.5%, putting it on track for its worst drop since President Donald Trump’s April 2025 “Liberation Day” tariffs. Every S&P 500 sector lost ground.

Technology shares drove much of the damage. Nvidia dropped 7.5% after falling 3.2% the day before. AMD sank 14.6%, while CoreWeave fell 12.1%. The Dow was down 1,433 points, or 2.7%, and the Nasdaq fell 4.7% at 12:26 p.m. Eastern time.

The pressure reflected worries that expensive borrowing and possible government restrictions could slow new data-center construction. That would reduce demand for chips, electricity, and related services. The losses overwhelmed strong results from AppLovin and Tapestry, whose shares rose sharply.

An AI data center is a facility filled with powerful computer servers that train AI systems and deliver AI services over the internet. Unlike ordinary office computing, AI workloads process enormous amounts of data and perform many calculations simultaneously. That requires specialized chips, especially the kind produced by Nvidia and AMD.

The chips work together across large groups of servers. More servers mean more chips, networking equipment, storage, and cooling systems. Each chip uses electricity while calculating, and the facility also needs power to remove the heat those calculations create. This is why AI data centers consume substantial electricity and purchase equipment and services on a large scale.

The article connects this buildout to broader economic growth. It says AI demand is vacuuming up chips, electricity, and other products and services. If construction slows, those purchases could fall, weakening companies tied to the boom.

Major AI-related companies suffered steep losses as investors questioned whether the data-center construction boom can continue at its previous pace. Nvidia fell 7.5%, making it the biggest drag on the S&P 500. AMD dropped even more, losing 14.6% during the session.

CoreWeave, which rents access to AI chips through the cloud, fell 12.1%. GE Vernova, a company helping power data centers, declined 6.1%. Nvidia had already fallen 3.2% the previous day, showing that the selling pressure was continuing rather than appearing in a single session.

These declines matter because AI-related firms have become large and influential parts of the market. The article links the weakness to worries about expensive borrowing and possible government pauses on new data centers. Slower construction could reduce future orders for chips, power, and related services.

Nvidia can move the S&P 500 unusually strongly because the index gives greater influence to companies with larger market values. Nvidia has grown into one of Wall Street’s largest stocks through the AI boom. Its size means its share-price changes affect the index more than similarly sized percentage moves in most other U.S. companies.

The article gives a clear example. A 1% move in Nvidia has more effect on the S&P 500 than a 1% move in nearly any other U.S. stock. Therefore, Nvidia’s 7.5% decline became the index’s heaviest individual drag during the sell-off.

This creates a feedback effect for headline market performance. When a very large company falls sharply, it can pull down the index even if some smaller companies rise. The article notes that AppLovin gained 28%, but that gain was overwhelmed by broad losses led by AI-linked giants.

Building an AI data center requires heavy upfront spending on land, buildings, servers, chips, power systems, and cooling. Companies often borrow to fund that investment. When borrowing costs rise, interest payments increase, raising the total cost of each project and reducing its expected return.

The article says AI-related stocks have been swinging because investors worry construction may slow as borrowing becomes more expensive. A company might delay, shrink, or cancel a planned facility if financing costs make the project less attractive. That decision would affect more than the builder itself.

Fewer new facilities would need fewer AI chips from companies such as Nvidia and AMD. They would also require less electricity and fewer products and services connected to the buildout. Because the AI boom is supporting much of U.S. economic growth, a slowdown could pressure many linked businesses and the wider economy.

The election results added a policy risk to the financial concerns already weighing on AI companies. The article says nationwide Democratic victories raised the possibility that governments could force pauses in the construction of additional data centers. Investors therefore had to consider not only project costs, but also whether projects would receive permission to proceed.

A government pause would interrupt the chain behind the AI buildout. Fewer new facilities would be built, so developers would buy fewer chips. Demand could also weaken for electricity, power equipment, and other products and services used by data centers. Nvidia, AMD, and GE Vernova are examples of companies exposed to different parts of that chain.

The concern is a possibility, not a confirmed nationwide construction halt. The article also notes that stock prices depend more over time on profits and interest rates than on who controls Washington. Those longer-term forces will help determine whether the election concern becomes lasting economic pressure.

Stock prices represent investors’ expectations about future company earnings, while interest rates affect the value and financing cost attached to those expectations. Stronger profits can support higher share prices. Higher rates can make borrowing more expensive and make safer investments, such as bonds, more attractive compared with stocks.

The article illustrates both forces. AppLovin jumped 28% after reporting better-than-expected revenue and profit and approving a stock-buyback program. Tapestry rose 6.6% after stronger profit and higher forecasts. At the same time, Treasury yields eased, suggesting investors were seeking safety as stocks fell.

Political events still matter, especially when they could change rules or spending. The election raised concerns about data-center pauses. But the article says markets depend much more on profits and interest rates over the long term than on who controls Washington. That makes company results and borrowing conditions crucial for future returns.