AI Is Erasing 16,000 U.S. Jobs Per Month — and Gen Z Is Absorbing the Worst of It
Goldman Sachs research finds AI eliminates ~25,000 U.S. jobs monthly while creating ~9,000 — a net loss of 16,000 per month, with Gen Z bearing disproportionate displacement in entry-level roles.
Why this matters
Goldman Sachs research finds AI eliminates ~25,000 U.S. jobs monthly while creating ~9,000 — a net loss of 16,000 per month, with Gen Z bearing disproportionate displacement in entry-level roles.
New Goldman Sachs research puts a number on AI-driven job displacement: approximately 16,000 net U.S. jobs eliminated per month. The underlying math — AI substitution removes about 25,000 jobs while AI augmentation adds back roughly 9,000 — reveals a labor market where AI is creating new roles, but not enough to offset the ones it's eliminating. And the workers absorbing the most displacement are Gen Z, the generation that entered the workforce expecting a different economy.
~25,000 jobs per month eliminated through direct AI substitution — AI systems replacing functions that humans previously performed
~9,000 jobs per month created through AI augmentation — new roles that emerge as humans work alongside AI, from AI trainers and prompt engineers to compliance and quality-oversight functions
The ratio — roughly 2.8 eliminations for every 1 creation — means the current period of AI adoption is a net negative for employment headcount, even as it creates genuinely new categories of work.
These are national estimates, not sector-specific figures. The displacement rate varies significantly by industry, with knowledge-work roles showing the highest substitution rates in the current phase.
Why Gen Z Is Hit Hardest
Gen Z workers — roughly those born between 1997 and 2012, with the workforce-age cohort primarily in their early-to-mid 20s — are concentrated in exactly the roles AI is automating most aggressively right now:
Data entry and processing — AI handles extraction, categorization, and entry at near-zero marginal cost
Customer service — agentic AI handles the majority of tier-1 and tier-2 contacts in companies that have deployed it
Legal support — document review, contract summarization, and legal research have been dramatically compressed by AI, reducing junior associate hours required
Billing and financial administration — AI processes invoices, reconciles accounts, and handles routine financial operations that were previously entry-level finance roles
These aren't coincidentally Gen Z jobs — they're entry-level roles. And entry-level roles are disproportionately routine and structured, which is exactly what current AI systems automate most effectively. Gen Z also lacks the accumulated expertise that insulates senior workers; without years of specialized knowledge and professional relationships, they have the least buffer against displacement.
Get this in your inbox.
Daily AI intelligence. Free. No spam.
The Career Trajectory Problem
The displacement pattern isn't just about whether Gen Z can find jobs today. It's about what happens to career development when entry-level positions disappear.
Entry-level roles in knowledge work have historically served a function beyond their immediate output: they're how junior workers build skills, develop professional relationships, and earn the credibility to move into more complex and autonomous work. A junior associate who spends three years doing document review learns contract law through that work. A junior analyst who spends two years building financial models learns how businesses actually work through the process.
When those roles are automated, the traditional apprenticeship pipeline breaks. Senior workers are insulated by accumulated expertise and relationships that took years to build. Gen Z workers are trying to build that foundation on ground that no longer exists in the same form.
This is qualitatively different from previous technology-driven job displacement. When ATMs replaced bank tellers, the affected workers were mid-career. When word processors replaced typists, the affected workers were concentrated in a specific support function. The current displacement is hitting the entry point for an entire generation's career trajectories, across multiple industries simultaneously.
What Can Actually Be Done
The honest answer is that the policy response to AI-driven displacement is lagging the displacement itself. Several approaches are being discussed with varying seriousness:
Apprenticeship and workforce training programs — reorienting public workforce development toward AI-adjacent skills rather than the roles AI is replacing
Curriculum reform in higher education — preparing students for a labor market that looks different from the one their professors graduated into (Stanford announced a $1 million initiative this week to do exactly this)
Labor market monitoring — building better real-time data on AI displacement rates to inform policy before the impacts compound further
Social insurance design — examining whether existing unemployment insurance and safety net programs are structured for the pace and pattern of AI-driven transitions
None of these are fast solutions. The 16,000 jobs per month figure is happening now, and policy timelines run in years.
What to Watch
Watch for the Goldman research to be updated quarterly — the 16,000 figure represents the current phase of AI adoption, and both the substitution and augmentation rates will shift as AI capabilities advance and as industries adapt. Watch also for Bureau of Labor Statistics data cuts that attempt to isolate AI-driven displacement from cyclical employment trends. Better measurement is the precondition for better policy, and right now the measurement is still catching up to the reality.
Hector Herrera is the founder of Hex AI Systems, where he builds AI-powered operations for mid-market businesses across 16 industries. He writes daily about how AI is reshaping business, government, and everyday life. 20+ years in technology. Houston, TX.
