AI Accelerates U.S. Labor Decline as Union Membership Hits Record Low

The share of workers belonging to unions has been dropping for decades. In 1983, 20.1% of U.S. employees were union members; by 2024 that figure had fallen to 9.9%, a historic low. Private‑sector unionization is even lower, at 5.9%. This trend reflects long‑standing policy choices such as at‑will employment rules, limited welfare programs, and deregulation that weaken collective bargaining. The current administration has accelerated these pressures by dismissing the head of the National Labor Relations Board and issuing orders that curb federal employee unions. The United States differs from many peer nations in that it lacks a strong safety net and allows employers to terminate workers without cause in most states. This legal backdrop makes it easier for firms to adjust headcount quickly when new technologies emerge.

The Department of Government Efficiency (DOGE) eliminated about 300,000 federal positions by March 2026, removing roughly 10% of the civilian workforce. DOGE used an AI‑based screening process: employees submitted written justifications for their roles, and an algorithm decided who stayed. Separately, a McKinsey survey from November 2025 found that 32% of companies anticipate AI will reduce their total staff by at least 3% within the next year. In the same survey, only a median of 17% reported actual AI‑related cuts in the prior year, indicating a growing expectation of future reductions.

These numbers show AI acting as an accelerant on a labor market already tilted toward efficiency over job security. While some firms have created new roles—13% hired AI compliance specialists and 6% added AI ethics specialists—the overall outlook points to net job losses, especially among low‑ and middle‑income workers who express the highest concern about AI’s impact. The pattern of cutting staff before AI tools are fully deployed suggests decisions are driven by cost‑saving goals rather than proven technological capability.