The concern sharpened at the Federal Reserve’s Sept. 15-16 meeting, where policymakers delivered their first rate increase in three years. Minutes released afterward show officials wrestling with an economy still buoyed by investment, credit and elevated stock prices even as energy costs and the AI buildout add fresh inflation risks.¹

The central fear was not simply that data centers, power demand and AI-related investment would raise prices in isolated corners of the economy. It was that those increases could spread. “A couple of participants” argued that a higher policy rate would stop price rises linked to energy disruptions and AI demand from “broadening out and generating more persistent inflation dynamics.”¹

That marks a notable shift in emphasis. Officials are increasingly identifying the artificial-intelligence buildout—not tariffs—as a key explanation for continued goods-price pressure, according to a separate report on the minutes.² The Fed also noted that the scale and speed of AI investment had repeatedly surprised to the upside, lifting investment and wages in some sectors while potentially delivering productivity gains later.

But the minutes reveal a split over the remedy. Several officials said policy was “not restrictive or only mildly restrictive,” a warning that rates may be doing too little to cool demand. Most participants thought another increase would probably be appropriate by year-end.¹

Since that meeting, however, the case for urgency has weakened among some senior voices. New York Fed President John Williams said there was “no need for urgency,” while Vice Chair Philip Jefferson said officials might need more time to assess whether another move is warranted.¹

The next inflation report will test both views: whether AI-fuelled expansion is becoming a durable price problem, or whether the Fed can afford to wait before tightening again.