AI jobs: Why the panic may be premature

2 Min Read

AI jobs: data versus doom

Recent headlines warn that AI jobs will vanish overnight, citing layoffs at Coinbase, Meta and Cisco. Yet the Bureau of Labor Statistics reveals that unemployment among occupations most exposed to AI is actually ↑ 0.4% lower than in less‑exposed roles, and there is no mass migration to manual‑labor sectors. Erika McEntarfer, former BLS chief, notes that only one in five firms uses AI for core functions, a reality check on the hype.

“Disruption is not yet here, and we have time to plan,”

she told Reuters. The broader market remains shaky for recent graduates, whose unemployment sits at ↓ 5.6%, a level not seen since the post‑pandemic surge. While entry‑level software and customer‑service roles for 22‑to‑25‑year‑olds have slipped, older workers in the same fields are still being hired, suggesting AI replaces codified knowledge more than tacit expertise. Stanford’s Digital Economy Lab, using ADP payroll data, flags a ↓ 16% dip in entry‑level positions in AI‑exposed occupations after 2024, but overall coder employment continues to rise, albeit slower. Wage growth in AI‑heavy sectors outpaces the average, implying firms still value experience that AI cannot mimic. The evidence points to a transition, not a collapse: businesses are augmenting workers, not eliminating them. As Harvard economist David Deming observes, we are “flying blind” without finer‑grained metrics, prompting a call for more granular surveys to track AI adoption across tasks. The takeaway: AI jobs are reshaping work, but the labor market’s shockwaves remain modest, buying policymakers time to craft reskilling programs before any sudden upheaval.

Dispatch from Zara Blackwood (Rapid Response Intelligence Analyst).