#沃什称AI是否引发通胀取决于美联储 I. Background: A Congressional Clash Over AI’s Inflation Traits

In July 2026, Federal Reserve Chair Kevin Wosch, at the Fed’s semiannual monetary policy hearing in Congress, for the first time publicly listed large-scale investment in the AI space as a core variable influencing inflation. Faced with lawmakers’ questions about whether the AI boom could push inflation higher, Wosch offered a thought-provoking answer: whether AI will cause inflation depends on the Federal Reserve.

While this statement seems to hand decision-making power to the central bank, it actually contains a deeper logic about how prices and inflation are distinguished, how the supply side responds, and how the central bank’s role is defined.

II. Wosch’s Core Argument: Price Changes ≠ Inflation

Wosch’s reasoning framework rests on three progressively layered judgments:

First, acknowledge that AI is driving up prices. Wosch said plainly, “The price spikes caused by AI are real, and I don’t want to downplay that.” AI infrastructure buildouts have already materially pushed up the prices of products such as computer chips, and are expected to further raise overall price data over the next 12 months.

Second, but rising prices do not equal inflation. Wosch’s core distinction is this: inflation is persistent and broad-based, a general rise in prices, while AI-driven price increases are a one-off supply-side phenomenon. “I don’t think a one-time price move will necessarily raise inflation, because the supply side will respond.”

Third, AI is fundamentally different from external conflicts. Wosch drew a key distinction between the inflation caused by geopolitical conflict and the price increases brought by AI investment—conflict weakens the economy’s capacity to supply, whereas AI investment raises prices while the supply side expands alongside it, ultimately forming a new equilibrium.

III. Original Analysis: Why “Depends on the Federal Reserve”?

When Wosch says whether AI triggers inflation “depends on the Federal Reserve,” the phrase can be interpreted in at least four layers:

1. The central bank’s judgment: distinguishing “noise” from “signal”

Wosch’s core point is that whether price changes evolve into inflation depends on how the central bank interprets and responds. If the central bank mistakenly reads a one-time surge in prices as a sign of persistent inflation and tightens monetary policy too aggressively, it could end up choking off the productivity gains brought by AI. If it instead views the move as a temporary supply-side adjustment and remains patient, it may leave room for AI’s longer-term disinflationary effects. As analysts have noted, Wosch is “willing to be patient with inflation sources he considers to be temporary.”

2. Managing expectations: the Fed’s self-fulfilling effect

Inflation itself has self-fulfilling properties—if the market expects inflation to rise, companies will raise prices in advance, and unions will demand higher wages, making inflation real. Wosch’s wording is essentially telling the market not to automatically equate the prices pushed up by AI investment with inflation. This expectation guidance in itself is an anti-inflation tool.

3. Time mismatch: demand arrives first, supply comes later

Wosch’s optimism rests on the assumption that the supply side will “respond.” The problem is the time mismatch—AI-driven demand shocks (data center construction, chip procurement) are happening now, while improvements in productivity and supply expansion take years to show up. The June FOMC minutes show that most officials believe the strong demand for AI infrastructure “could lead to more persistent inflation pressure.” In this time window where the demand effect has arrived but the supply effect has not, every central bank decision matters for the direction of the inflation path.

4. Institutional level: the Fed is reshaping its judgment framework

Wosch is not just speaking in generalities. He announced the formation of five working groups, and the “productivity and employment” group for the first time includes people from Silicon Valley industries in the Fed’s decision-making circle to specifically assess the economic impact of general-purpose technologies such as AI; another working group is tasked with studying how to “obtain more accurate data from external sources” to improve inflation measurement. This shows that “depends on the Federal Reserve” is not merely a talking point, but the launch of an entire institutional build-out.

IV. The Tension Behind the Contradiction: Wosch’s Optimism vs. Colleagues’ Concerns

Wosch’s stance is not a consensus within the Fed. The June FOMC minutes show that “most participants” believe the strong momentum in AI business spending “could lead to more persistent inflation pressure.” New York Fed President Williams went even further, saying that AI-driven demand has become his “main inflation concern.” This internal disagreement actually confirms Wosch’s judgment that “depends on the Federal Reserve”—with the same facts, different interpretations will lead to different policy paths.

Wosch’s statement that “whether AI triggers inflation depends on the Federal Reserve” is, in essence, about securing greater discretion for the Fed’s judgments and policy flexibility. It is both an expectation-guiding move for the market and a restructuring declaration for the internal policy framework. In a mismatched period where the AI demand effect has arrived but the supply effect has not, how the Fed distinguishes “one-off price adjustments” from “persistent inflation pressure” will determine whether the U.S. economy can embrace AI’s dividends while keeping inflation from getting out of control. This is not only an economics issue—it is a race over cognition and time.