A slight majority of economists anticipate that AI will weigh on US employment in the year

In Indeed's quarterly Hiring Lab survey, 52% of the 120 specialists surveyed expect at least a slight brake on employment, and only 4% expect a major productivity shock.

In July 2026, the Hiring Lab, Indeed's labor market analysis lab, surveyed one hundred and twenty economists and specialists from academia, financial institutions, and independent research centers. The central question was the net effect of AI on U.S. employment in the next twelve months.

The response is measured rather than alarmist: 52% of respondents expect at least a slight brake, 35% a net gain and 13% no perceptible effect. The consensus is therefore for a slowdown, not a collapse.

Productivity, a promise postponed

The most telling contrast is in the expected gains. Almost all respondents believe that AI will improve productivity within three years, but 70% expect only a modest to moderate improvement, and only 4% expect a transformative effect. Unemployment forecasts are following the same gentle trend: 4.29% in July 2026, 4.37% expected in December, 4.38% in June 2027. The gap with the rhetoric about the massive replacement of white-collar workers remains, for the time being, wide.