Financial markets are currently heavily focused on geopolitical risks and the immediate inflationary impact of rising energy prices. However, a growing consensus suggests that core US inflation dynamics are reverting to their pre-COVID patterns. When considering the rapidly increasing probability of significant labor market disruption due to generative and agentic AI, the conclusion for the inflation market might be unsettling. The article posits that markets should likely be pricing a negative core inflation risk premium beyond the near term, rather than a positive one, marking a significant shift from earlier expectations.
Before the COVID-19 pandemic, the US core CPI ecosystem was remarkably stable, with core goods inflation around 0%, rent inflation at 3.0-3.5%, and core services excluding rents falling between these figures, leading to an average of about 2%. This distribution anchored policy and market pricing for years. The current landscape indicates that the bar for supercore inflation is now higher, yet wage growth has already decelerated to pre-COVID levels. The Atlanta Fed wage tracker and Barclays’ state-space decomposition suggest wage momentum is at or below the level needed to achieve 2% inflation in this new regime, given softer rents and flat goods prices.
Generative and agentic AI are expected to amplify this trend. The article warns against dismissing AI's impact on the grounds that labor data doesn't yet show stress, comparing it to ignoring hurricane forecasts because there's no current flooding. It suggests that by the time labor market stress becomes evident, the repricing will have already occurred. If wage growth is already considered too soft for the new core inflation arithmetic, accelerating productivity-enhancing or labor-substituting technologies from AI could further depress wages. If tariff pass-through fades and core goods return to 0% in the second half of 2026, and rent inflation runs below its pre-COVID trajectory, then core services excluding rents would need to run higher than pre-COVID levels just to keep the overall core basket near 2%.
While the primary article emphasizes the disinflationary potential of AI through labor market disruption, it also briefly acknowledges counterpoints. Geopolitical tensions, like an energy price shock, can drive front-end inflation higher. Additionally, massive AI capital expenditure could be supportive of inflation pressure, and electricity prices are broadly rising. However, the article's core argument leans heavily on AI's potential to soften labor demand and thereby lower inflation, describing a scenario where wage growth undershoots just as the supercore inflation hurdle has increased.
Separately, other sources indicate that the immediate impact of AI is actually inflationary due to high demand for memory chips, which have seen a record 14.5% increase in software and computer accessories prices and a 27% increase in electronic components for producers in May. Bloomberg Economics calculates that the memory squeeze alone will add 0.4 percentage points to headline inflation. Experts like Torsten Slok from Apollo Global Management foresee the initial AI boom as inflationary, citing semiconductor and energy prices, and labor. This short-term inflationary pressure from AI infrastructure costs contrasts with the longer-term disinflationary pressure discussed in the main article, creating a complex picture for central bankers.
