The current rise in the US stock market is highly reliant on the AI infrastructure chain, and the market is unprepared for policy risks.

Authored by: Li Jia

Source: Wall Street Journal

The AI investment boom that has propelled the US stock market to new highs is facing increasing social and political resistance.

Reports indicate that as the surge in electricity consumption by data centers drives up residential electricity bills, and concerns about AI replacing jobs continue to grow, opposition to AI infrastructure is rapidly spreading from public opinion to the legislative level in many parts of the United States. Virginia was the first to pass a tax bill on electricity used by data centers, and the New York state legislature is pushing a ban on the construction of large data centers into the governor's approval process, marking the transformation of "anti-AI sentiment" into actual policy risks.

This change has raised alarms on Wall Street. The core drivers behind the S&P 500 index's rise this year have almost all stemmed from the AI infrastructure chain. Evercore ISI warns that the spread of the anti-AI wave is far outpacing market expectations, and once political figures begin to respond to voter pressures, investors will have to reassess the regulatory and tax risks within their AI investment logic.

For technology stocks that are at historical highs, a previously overlooked issue is becoming increasingly important: Is AI experiencing its own "backlash"?

AI infrastructure surge, US stocks highly dependent on a single narrative

Over the past two years, AI infrastructure development has become the strongest growth story in global capital markets, and this trend is expected to strengthen further in 2026.

From storage chip manufacturers Micron Technology and Western Digital to companies like Caterpillar that provide power generation equipment and engineering machinery for data centers, a large number of enterprises are benefiting from the surge in AI capital expenditures. Data shows that this year, eight core stocks related to AI infrastructure have contributed nearly two-thirds of the 7.5% increase in the S&P 500 index.

However, this highly concentrated rise also means the market is becoming increasingly fragile.

The Philadelphia Semiconductor Index has seen a cumulative increase of nearly 90% this year, potentially marking its best annual performance since 1999. However, just this week, due to reports from South Korean media that SK Hynix is slowing down its AI memory chip expansion plans, the index plummeted 7.9% in a single day, indicating that the market has become exceptionally sensitive to the outlook for AI demand.

Although Micron subsequently released a sales forecast that exceeded expectations, momentarily stabilizing market sentiment, investors are beginning to realize that the risks facing AI trends may not solely stem from corporate profitability.

From electricity bills to employment: "Anti-AI" sentiment begins to enter the legislative process

The most direct trigger for the current anti-AI wave stems from energy issues.

As the demand for electricity in large data centers surges, residents in many parts of the United States are facing rising electricity bills, leading to growing dissatisfaction among local governments and voters. Meanwhile, worries about AI replacing jobs are also starting to spread to a broader demographic.

This sentiment is rapidly transforming into policy action.

From California to Georgia, many local governments have suspended approvals for certain data center projects; Maine Governor Janet Mills vetoed a statewide freeze on data center construction this year; the New York state legislature has passed a one-year ban on the construction of large data centers, pending a final decision from Governor Kathy Hochul.

More symbolically, Virginia, home to the largest concentration of data centers globally, passed a bill this week imposing a tax on electricity used by data centers, seen as the first case in the United States to implement a special tax on AI infrastructure.

At the federal level, some progressive lawmakers have begun discussing the possibility of taxing AI companies and limiting the expansion of data centers.

Wall Street begins to worry about a "policy turning point"

For investors, what really deserves attention is not the individual state policy actions but the potential for changes in the political landscape.

Gabelli Utilities Fund portfolio manager Tim Winter pointed out that there are 36 gubernatorial elections happening across the United States this year, and in the face of issues like rising electricity bills, public utility regulators and state governments may gradually change their stance on data center expansion.

This concern has already begun to affect the pricing of related sectors. In April of this year, the stock prices of several Pennsylvania utility companies fell due to Jefferies’ warning about the "affordability risk of electricity."

Evercore ISI believes that what is more concerning is the rapid pace at which this discussion is advancing. If voter dissatisfaction with AI continues to accumulate, politicians may eventually have to take action in response, which could challenge the long-held market assumption of "regulatory leniency."

The biggest risk to the AI trend may not come from technology

Evercore ISI further points out that in the coming years, investors need to pay attention to more than just tax issues.

The government may require new AI models to undergo mandatory reviews, and courts may issue rulings unfavorable to model developers regarding copyright infringement or liability for AI-generated content. These regulatory and legal risks have previously been largely excluded from market pricing but are gradually becoming critical variables influencing the development of the AI industry.

For the US stock market, which has heavily relied on AI narratives for support, the greatest challenge may not be a slowdown in technological advancement but how political forces intervene when the conflict between the wealth generated by AI and the societal costs it brings begins to intensify. At that point, Wall Street will need to answer a question again: Is AI truly a technological revolution, or is it the starting point of the next regulatory storm?