AI is cutting 16,000 U.S. jobs a month—and Gen Z is taking the brunt, Goldman Sachs says

Goldman Sachs finds AI erased roughly 16,000 net U.S. jobs per month over the past year (about 25,000 jobs/month lost to substitution vs ~9,000/month gained via augmentation). The report flags outsized pain for Gen Z/entry-level workers and estimates a one standard-deviation rise in AI substitution exposure widens the entry-level-to-experienced wage gap by ~3.3 percentage points. Goldman cautions the net aggregate impact is likely smaller because the analysis understates offsetting hiring tied to AI infrastructure (data centers, power, construction) and demand created by productivity-driven market expansion.

Analysis

AI-driven substitution is already reshaping labor composition in a way that will depress early-career earnings and cyclical consumption for years. Younger cohorts concentrated in routine white‑collar roles will see slower wage progression and delayed credit/housing milestones, which implies a multi-year drag on lower‑end discretionary spending and starter‑home demand unless retraining or new hiring pathways scale quickly.

Corporate response will bifurcate: firms with high volumes of low-skilled admin work will accelerate automation and reclassify headcount toward contract, gig, or offshore models, while firms that rely on specialist judgment will bid up senior talent, tightening margins in professional services and legal sectors. That compression creates second‑order winners in tooling and cloud infra (where recurring SaaS and data-center revenue accelerates) and losers among legacy staffing and BPO providers that depend on volume hiring.

On the supply chain side, the labor displacement signal should translate into sustained capex for data centers, power delivery, and chip fab equipment, with deployment and hiring cycles lagging software adoption by 6–18 months. This timing mismatch creates a window where productivity gains reduce labor demand even as construction and hardware hiring ramps — a stagflationary microshock for certain local labor markets.

Key catalysts that could reverse or amplify these trends include tougher labor/regulatory interventions or subsidized retraining (which would blunt net job losses) versus rapid generative model improvements and commoditization of AI tooling (which would accelerate substitution). Monitor capex guidance from hyperscalers and quarterly hiring mix disclosures as high‑frequency indicators of the infrastructure offset arriving or failing to materialize.

Key Decisions for Investors

• Pair trade (6–18 months): Long data center REITs EQIX (Equinix) and DLR (Digital Realty) vs short staffing/contingent labor names MAN (ManpowerGroup) and ASGN. Rationale: infra capex and recurring colocation revenue should outpace volume hiring; target 20–30% upside on longs with 15–20% downside protection via stops on a slowing capex print.
• Long semiconductor equipment exposure (12–24 months): LRCX or AMAT on the view that fab and packaging spend expands as firms onshore AI supply chains. Use buy‑write or long‑dated calls to limit downside; reward skew if equipment orders reaccelerate—expect 30–50% upside if cycle firms up, but material downside if hyperscaler capex is deferred.
• Long cloud/AI platform leaders (9–12 months): MSFT or GOOGL secular play to capture SaaS+compute consumption even as headcount shifts. Prefer covered-call or call-spread to finance carry; downside risk concentrated in macro slowdown or regulatory curbs on model training.
• Event hedge (3–12 months): Buy protection or short small-cap staffing and BPO ETNs ahead of major earnings where companies report entry-level hiring metrics. Catalyst: earlier-than-expected declines in entry-level hiring would compress multiples; reward asymmetric as staffing firms are high beta to entry‑level demand.