The AI ​​bubble worries Wall Street: the Nvidia case and the risks for the USA and Europe

Artificial intelligence continues to attract huge investments, but on Wall Street the enthusiasm of recent years is giving way to some concerns. Technology companies spend billions on data centers, while investors question the sustainability of this growth. What is causing discussion above all is Nvidia, the American chip giant, which in the last month has lost over a quarter of its value on the stock market.

Complicating the picture was the resignation of founder Jensen Huang, involved in an affair that raised questions about company security. Behind the reductions, however, there is also the doubt that the high costs of the infrastructure necessary for the development of this technology will not be repaid. The situation has reignited debate about a possible AI financial bubble, with repercussions that could go beyond the United States.

Why Nvidia is losing value on the stock market and what does the Huang case have to do with it?

On November 12, 2026, Nvidia announced that Jensen Huang resigned as CEO and from the board of directors. Colette Kress, former financial director, was called to temporarily lead the company.

The resignation came amid reports of the manager’s personal relationship with a woman who, according to the American press, is suspected of having links to Chinese intelligence. Already on November 8, Nvidia’s board of directors announced an independent investigation, specifying that it had no evidence to believe that corporate secrets or national security had been compromised.

For investors, what matters most is the exit of a manager who led Nvidia for over thirty years, building relationships with customers and governments. Although the company has ensured continuity in projects, the future of its leadership remains uncertain. According to rumors, Huang is considering a return as CEO and has spoken about it with some people close to him. At the moment, however, no official confirmation has arrived.

On November 12, Nvidia shares closed at $166.19, down 4.34%. On Friday the 13th they lost a further 2%, falling to 162.85 dollars. The decline, however, had begun before the resignation: on 6 November the stock was already worth around 17.5% less than on 12 October. By Friday’s close, the loss from that date had risen to about 28%. Also weighing were fears that large technology companies could slow down the construction of data centers and, consequently, purchases of processors.

The difficulties did not remain limited to Nvidia. On Friday 13 November the Nasdaq Composite, a US index strongly influenced by technology stocks, lost 0.91%. In the same session, AMD and Broadcom lost 8.8% and 4.8% respectively. However, the declines did not uniformly affect Wall Street: the S&P 500 fell by 0.40%, while the Dow Jones gained 0.10%.

Why American markets fear an AI bubble

Nvidia designs GPUs, processors capable of performing many calculations simultaneously, widely used to train and operate numerous AI systems. Among the main customers are companies that manage large data processing centers, using the computing power for their own services or making it available to other companies.

In the quarter ending July 2026, the data center business generated $89 billion of Nvidia’s $96.2 billion in overall revenue, or more than 92% of revenue. Total revenue had more than doubled compared to the same period last year. The problem, therefore, is not a collapse in sales, but the fear that such rapid growth cannot continue at the same pace.

Building and managing a computer center requires enormous expenses for processors, cooling systems and electricity. They are costs to be borne without always having the certainty that future revenues will be sufficient to cover them. Some operators use loans to finance these projects, but if revenues are lower than expected or credit becomes more expensive, paying the money back can be difficult. New facilities could thus be postponed or scaled back, impacting chip orders.

Another aspect concerns the financial relationships between companies in the sector. Nvidia also invests in some companies that buy or use its technologies. In January 2026, it announced a $2 billion investment in CoreWeave, a company that makes the computing power of its data centers available to other businesses. In March it announced an operation of the same amount in Nebius, specialized in artificial intelligence infrastructures.

These agreements have fueled the debate on the so-called circular financing, a mechanism in which the capital invested by a company can contribute, directly or indirectly, to supporting the purchases of its own products. It doesn’t mean that the money automatically goes back to whoever provided it: CoreWeave clarified that the funds received from Nvidia would not be used for the direct purchase of its processors, but for the expansion of its activities.

The risk is rather that of creating economic ties that can amplify the difficulties. If a data center operator loses customers or fails to finance new facilities, it may order fewer chips. At the same time, the value of investments made by its partners may decline. A problem initially limited to one company would thus end up involving others.

The International Monetary Fund recently flagged the risks of a correction in AI-related stocks, exacerbated by indebtedness and financial relationships between companies in the sector. Nvidia’s accounts confirm that demand for processors remains high, but they do not demonstrate that all the services built thanks to these infrastructures are profitable. The question is whether future earnings will be able to justify the valuations achieved on the stock market.

What are the risks for the United States and the rest of the world

American tensions are also felt abroad, but not all stock markets react in the same way. On the morning of Monday 16 November, Asian markets recorded new declines, particularly marked in Taiwan and South Korea, while Hong Kong moved against the trend. In Europe, on Friday 13 November, the STOXX 600 index closed essentially unchanged, but the technology sector lost 1.3%. The difficulties of the AI ​​sector, therefore, do not necessarily translate into a uniform decline in all markets.

Debate over a possible AI bubble had already emerged in China, where the decline in the STAR 50 technology index had fueled investor concerns. The two cases have different characteristics and the Chinese declines do not allow us to predict what will happen in the United States.

The artificial intelligence industry involves an international network of companies. Semiconductors are designed, manufactured and assembled in different countries, while building a data center requires electrical equipment and components supplied by companies all over the world. If US investment slows for a long time, these suppliers may also receive fewer orders, despite not directly developing AI systems.

Savers can also be affected: American technology giants are present in numerous investment funds and ETFs, funds traded on the stock exchange. A significant drop in stocks could therefore reduce the value of some portfolios, even for those who do not directly own Nvidia shares.

Above all, it is important to distinguish a correction in stock market prices from a crisis in the real economy. Falling stock prices don’t automatically mean companies produce less or cut jobs. The problem would become more serious if financial difficulties lead to canceling projects, reducing hiring or interrupting already planned investments. In this case, the consequences would no longer only concern investors, but also the companies and workers involved in the artificial intelligence supply chain.