Meta spends $145B, cuts 8,000 jobs same day

Q1 Earnings Lock In the CAPEX-Headcount Trade-Off

• Meta raised its full-year AI capex guide to $125–$145 billion — explicitly tying the figure to the upcoming 8,000-job cut starting May 20; Zuckerberg told staff on the April 29 earnings call the cuts are needed to "offset other investments we're making," and Chief People Officer Janelle Gale said the company "cannot rule out further layoffs."
• Google Cloud grew 63% YoY to $20B; AWS grew 28% to $37.6B — both exceeded estimates, with Alphabet CEO Pichai saying the company is "compute constrained in the near term"; combined 2026 AI capex across five hyperscalers is on track to exceed $650 billion.
• Despite the efficiency rhetoric — Meta and Amazon executives collectively used the word "efficiency" 15 times across Q1 calls, and Microsoft's CFO said headcount would decline this year — a Washington Post analysis published May 1 finds tech giants have not significantly shrunk their overall workforces; the story remains hiring freezes and capex redirects rather than mass terminations.

---

Agentic AI and the "Big Freeze" Closing the Entry-Level Door

• A Yale CELI analysis published in Fortune identifies agentic AI — not generative AI — as the real labor disruptor: it shifts automation from discrete tasks to entire workflows, enabling firms to extract more output without backfilling departures; McKinsey survey data cited in the same piece shows 43% of companies expect AI not to change workforce size and 32% expect a 3%+ decrease achievable through attrition alone at typical 13% voluntary turnover rates.
• C.H. Robinson is handling 29% more LTL freight volume with 30% fewer employees than 2019, with ~50% of carrier bookings now agent-generated; real estate firms cut on-property labor hours 30% and headcount 15% via agentic systems, per the same Yale research.
• Entry-level software engineering postings are down 53% on Indeed; Verizon CEO Dan Schulman predicts a 30% rise in unemployment within 2–5 years if current agentic adoption trends continue.
• 4.2% of full-time early-career jobs on Handshake now require AI skills — nearly double from a year ago; 10.3% of internships mention AI, with the sharpest growth in government, healthcare, and education from near-zero; 62% of college seniors feel pessimistic about their careers, up from 46% in 2024, and total job postings on the platform are down 12% from pre-pandemic levels.
• Cognizant is running the bifurcated hiring playbook at scale: "Project Leap" cuts ~4,000 mid-level roles while simultaneously hiring 20,000+ freshers, explicitly moving away from people-heavy delivery models toward automation and AI-driven services; Oracle, TCS, Infosys, and HCL all reported layoffs or weak growth in the same period.

---

China Court Rules AI Replacement Cannot Legally Justify Termination

• The Hangzhou Intermediate People's Court ruled that an AI company acted unlawfully when it dismissed a quality assurance supervisor after LLMs took over his tasks — holding that AI replacement does not constitute a "major change in objective circumstances" under China's Labor Contract Law, and that offering the worker a role with a 40% pay cut does not qualify as a "reasonable" reassignment.
• A separate Beijing ruling in a map-data case found the risks of an employer's voluntary AI adoption "should not be forced solely onto employees" — establishing a second precedent that companies benefit from AI efficiency but cannot shift its costs onto labor.
• The rulings arrive as China's core AI industry exceeded 1.2 trillion yuan in 2025 across 6,200+ enterprises; AI agent penetration is projected to exceed 90% by 2030, making these labor protections increasingly consequential.
• In the U.S., state-level legal exposure is also proliferating: California's FEHA now prohibits AI-driven discrimination in termination and requires documented anti-bias testing; New York's WARN Act now requires disclosure when AI is a factor in layoffs; healthcare employers face compounded liability if AI-linked workforce reductions lead to declining quality-of-care metrics.

---

Goldman Softens Net Job Forecast; Small Business and Trades Absorb What Big Tech Drops

• A new Goldman Sachs report — softer than its earlier net -16,000 jobs/month figure reported in April — now projects AI could displace 1–4 million U.S. jobs annually but create more jobs than it destroys through efficiency-driven economic growth and new AI-focused roles; the bank does not anticipate a "job apocalypse" and notes current employment impact remains limited outside software development and specific white-collar sectors.
• Apollo Global's chief economist Torsten Slok argues AI efficiency in professional services will expand total demand — a "Jevons Employment Effect" — with Salesforce actively hiring 1,000 entry-level workers for AI systems and IBM tripling Gen Z entry-level hiring as counter-evidence to displacement narratives.
• Small businesses will hire an estimated 974,000 recent graduates in 2026, up from 962,000 in 2025, per a Gusto report; net new grad job creation has climbed from a low of 60,000 in 2023 to over 100,000 in 2026; the fastest-growing new grad roles are AI engineers (tech-native) and field managers and service technicians (AI-resistant trades).
• Blue-collar Gen Z enrollment in vocational colleges jumped 16% in 2024; construction trades enrollment surged 23% from 2022–2023, per the same Gusto data; workers aged 50+ face the opposite trajectory — structured, data-driven layoff models score cost-to-company and distance from revenue generation, hitting high-salary mid-career workers hardest; 44% of companies have already reduced management layers, and 1 in 4 workers aged 50–65 never return to work after a layoff.