U‑Michigan forecasters: AI investment bolstering growth but rising core inflation and oil shocks heighten risks

Yinuo Zhang, a senior economist at RSQE, and Gabe Ehrlich of the University of Michigan presented the joint U.S. and Michigan economic outlook at the May consensus revenue estimating conference. They said recent quarterly U.S. real GDP has been roughly 2% and that strong AI‑related capital spending has driven unusually large investment in computers and peripheral equipment, much of which appears to be imported.

The presenters emphasized inflation as a central near‑term risk. Zhang said core PCE inflation has accelerated recently and “several FOMC members prefer to stop signaling rate cuts,” noting that energy‑driven price spikes have pushed headline CPI higher. She flagged that some computer‑software and accessory price gains linked to AI demand have contributed to divergence between core PCE and CPI readings.

On the labor market, the presenters described a mixed picture: payroll job gains have averaged tens of thousands per month recently, while broader household employment and labor force participation have eased. Ehrlich said Michigan payroll jobs are expected to be modestly higher over the next two years even though calendar‑year 2026 will reflect some job losses already realized.

Both presenters said the largest immediate downside risk is higher oil prices driven by geopolitical tensions in the Middle East; that shock could feed into core inflation and reduce consumption. They added that the long‑term contribution of AI to productivity is uncertain because official data and input‑output measurement may lag the effects of imported hardware investment.

The forecasters projected U.S. GDP to slow modestly in 2027 after a pickup in 2026 and expected Michigan employment to recover only gradually, with unemployment near 4.5–5% over the forecast horizon. They underscored that incoming data in the months ahead will determine whether recent inflationary pressures persist.