McKinsey Report Projects 36 Million US Jobs Displaced by 2035 Amid AI Boom

Key points

• McKinsey estimates 36 million jobs' worth of labor demand will decline by 2035 due to automation, though 40 million new jobs are expected to be created.
• Office and administrative support roles face the highest impact, with 81% of work hours affected and a 38% reduction in labor demand.
• Approximately 11 million workers must change occupations, but nearly 45% face 'unpaved' pathways requiring costly reskilling or relocation.
• Paul Krugman argues that $10 trillion in AI data center spending will crowd out other investments, potentially reducing labor demand by 10 million jobs.
• McKinsey’s Eric Kutcher notes that while individual worker productivity has risen, enterprise-level productivity gains remain limited outside software development.

Background

This report follows earlier Bureau of Labor Statistics projections from September 2026, which forecasted modest overall job growth of 5.9 million by 2035, driven by healthcare and technology sectors. While BLS data highlighted declines in roles like word processors and data-entry keyers, the new McKinsey analysis provides a broader view of labor demand shifts, emphasizing that automation may 'reinvent' rather than eliminate many roles, such as in the legal field where AI handles research but human judgment remains essential.

How outlets are covering it

Outlets present diverging views on AI's economic impact. CNBC highlights McKinsey's optimistic view that new job creation will outpace declines, though it acknowledges the friction for displaced workers. Paul Krugman offers a starkly pessimistic take, arguing that AI infrastructure spending crowds out other capital investments, potentially reducing labor demand by 10 million jobs and worsening income inequality. Business Insider reports that while AI boosts individual productivity, it has not yet translated into broad enterprise-level gains, with only 37% of organizations seeing profit contributions from AI. The Financial Times headline suggests skepticism among economists regarding Silicon Valley's fears of mass job loss, though the full article is paywalled.

Why it matters

The findings underscore the urgent need for a 'system-level response' from employers and governments to support worker reskilling. With 770,000 workers needing to switch occupations annually, the transition rate is 3.6 times the historical average. Failure to address this gap could lead to significant wage suppression and economic instability, as AI investment may not generate the broad productivity gains currently promised by corporate leaders.

What to watch

Policymakers and employers must invest in reskilling programs to help the 11 million displaced workers transition into new roles. Companies are beginning to move beyond AI experimentation into 'scaling' modes, rethinking entire business processes to integrate AI more effectively. Workers are advised to focus on developing adaptable skills like problem-solving and critical thinking to maintain resilience in a rapidly changing labor market.