In March, the U.S. labor market added 178,000 jobs, a pace little changed from the prior month, according to the Bureau of Labor Statistics. That sluggishness is unfolding amid shifting White House policy, higher energy costs tied to conflicts in the Middle East, and growing evidence that AI is reshaping hiring and work.
Tech boosters argue AI and large language models will spark an economic boom by improving efficiency. But as AI is woven into everyday business processes, a gap is widening between that promise and the labor-market reality.
AI and slower hiring
Some high-profile figures have downplayed AI’s threat to jobs. Venture capitalist Marc Andreessen, for example, suggested fears about AI-driven displacement are exaggerated and pointed to reporting that tech job openings have increased. Business Insider cited TrueUp data showing tech job listings doubled to about 67,000 since 2023.
Open listings, however, don’t guarantee hires. BLS data show most new jobs in March were outside core tech: healthcare added about 76,000 positions, construction grew by roughly 26,000, transportation and warehousing added about 21,000, and social assistance increased by 14,000. Components of the tech sector tell a different story: computing infrastructure providers shed about 1,500 jobs, web search portals were roughly flat, and computer systems design and related services lost about 13,000 positions.
Analysts who study AI’s employment effects report sharper impacts. A Goldman Sachs analysis cited by Fortune estimated AI-related losses at roughly 16,000 jobs per month over the past year, with particularly steep declines in entry-level hiring. A 2025 SignalFire study found new-graduate hiring in tech is down about 50% compared with pre-pandemic levels, noting that smaller funding rounds, shrinking teams, and the rise of AI have closed what was once an open door for recent graduates.
Goldman Sachs warns these displacements could have persistent effects. Workers pushed out by automation often move into more routine occupations that demand fewer analytical and interpersonal skills, a shift that can depress earnings and labor-market prospects for years.
C-suite optimism vs. frontline experience
Executive enthusiasm for AI remains high. Harvard Business Review reports roughly 80% of senior leaders use AI weekly, and 74% say early deployments are yielding positive returns. But that optimism doesn’t always reach rank-and-file employees.
Surveys and reports show mounting workplace friction. Mercer found 43% of workers say their jobs have become more frustrating since AI tools were introduced. Workday’s research notes that for every 10 hours of efficiency supposedly gained from AI, nearly four hours are spent correcting errors or reworking outputs. HBR researchers coined the term “workslop” to describe superficially polished AI-generated content that lacks substance and pushes cognitive labor onto coworkers; they found 41% of employees have encountered such output, costing nearly two hours of rework per instance and harming productivity and trust.
Only a small share of organizations report consistent net-positive outcomes from AI: Workday’s respondents put that figure at about 14%. One reason for the divide between leaders and frontline staff is how each group uses AI. Senior leaders tend to apply AI for high-level synthesis, strategic drafting, and decision support—areas where models perform relatively well. Daily operational work, with long-established workflows, mixed technical comfort across teams, and high demands for accuracy, exposes AI’s limitations. When AI tools fail, the burden of cleanup often falls on nonexecutive employees.
Industry voices warn of an “AI tax”: more checking, more rework, heightened anxiety, and eroded trust. That friction undercuts the productivity gains many executives expect.
Policy conversations and responses
Some AI companies acknowledge these risks. OpenAI has publicly outlined policy proposals—framed as exploratory ideas—to help manage displacement, including expanded healthcare coverage, strengthened retirement savings, and a refreshed industrial policy agenda. The company warned that without policy keeping pace with technological change, social institutions and safety nets could lag behind.
That stance contrasts with more dismissive public comments from some tech leaders, and it highlights a growing debate: how to balance AI-driven efficiency gains with protections for workers and investments in reskilling.
Editorial note
This piece is produced in the style of long-form reporting and analysis. It does not constitute financial, legal, or investment advice. Editorial content reflects the authors’ reporting and analysis and is independent of advertisers and commercial partners.