AI, Inflation and Interest Rates

AI is currently acting as a mild inflationary force in the short term—driving up electricity demand, construction wages, and chip prices—despite its long-term potential to revolutionize productivity and reduce inflation. David Kelly of JPMorgan Asset Management argues that while Kevin Warsh’s vision of AI as a disinflationary catalyst is compelling, it’s premature to justify immediate Fed rate cuts. The immediate inflationary pressures from AI infrastructure spending, combined with supply-side constraints like reduced immigration and geopolitical tensions, suggest that headline inflation could reach 3.9% by May 2026. Yet, in the long run, AI’s productivity gains—amplified by declining union power, rising inequality, and enhanced price competition—could suppress inflation. However, the Fed cannot fix structural issues like housing affordability or inequality through monetary policy alone, as low rates in the past fueled asset bubbles and worsened inequality. The real takeaway: AI’s economic impact is asymmetric—short-term inflationary, long-term deflationary—and the Fed must resist the temptation to ease prematurely. The episode underscores a critical tension: while AI promises transformative efficiency, its current deployment is inflating costs across energy, labor, and materials. Even as workers fear job displacement, wage growth has slowed, signaling a labor market 'scare effect' rather than a tight labor market. Meanwhile, adoption is accelerating—50% of U.S.