‘AI-washing’ and ‘forever layoffs’: Why companies keep cutting jobs, even amid rising profits

While AI fuels revenue gains at leading tech suppliers, the technology so far explains a small share of 2025 U.S. job cuts—just 55,000 of 1.2 million announced (4.5%), according to Challenger, Grey & Christmas—raising concerns about “AI-washing.” ASML, which posted a 16% increase in net sales and 19% higher gross profit year-over-year, cut 1,700 jobs to trim bureaucracy, and Amazon—whose AWS grew 24% in the quarter—announced roughly 30,000 layoffs across two rounds even as it plans roughly $200 billion in capex (up ~60% year-over-year). The dynamic of profitable companies trimming headcount is creating morale and sentiment risks that could affect productivity and investor perception despite solid fundamentals.

Analysis

Market structure: The immediate winners are semiconductor capital-equipment incumbents (ASML) and datacenter supply chains (power, copper, server OEMs) as hyperscalers accelerate AI capex; ASML’s EUV monopoly gives >10%+ gross-margin runway and pricing power into FY26 given 16% sales / 19% gross-profit growth. Losers include parts of consumer-facing tech (Amazon’s margins) and high‑growth, low‑profit cloud peers where sustained capex (Amazon guiding ~$200bn) compresses near-term free cash flow and equity returns. Cross-asset: stronger capex increases commodity demand (copper, aluminum), pressures IG credit if capex funds leverage, and can lift USD if funded by overseas flows; expect elevated tech equity IV around earnings and option skew on AMZN/ASML.

Risk assessment: Tail risks include export/regulatory curbs on ASML to China (low probability, very high impact), an AWS capex overspend leading to credit-rating pressure at AMZN, or a macro slowdown that collapses datacenter demand; probability window 6–24 months. Immediate (days) risks are earnings/guide reactions; short-term (1–6 months) are orderbook/capex revisions; long-term (12–36 months) is structural labor displacement and demand reallocation. Hidden dependencies: workforce morale from “forever layoffs” can reduce innovation velocity (second-order hit to long-term revenue) and consumer demand if layoffs broaden.

Trade implications: Tactical long ASML (small, concentrated) and defensive short/hedge on AMZN are warranted; options can limit cost — e.g., 3–6 month AMZN put spreads sized to 1–2% portfolio. Pair trade long ASML vs short AMZN (notional balanced) captures structural divergence while neutralizing beta. Rotate 3–5% from consumer discretionary into semicap suppliers and copper/mining over the next 30–90 days; use stop-losses (10–12%) and rebalance on orderbook or AWS guidance triggers.

Contrarian angles: Consensus fears a systemic AI job apocalypse; data show only ~4.5% of announced cuts cite AI—market may be over-discounting consumer demand loss while under-discounting order backlogs for semicap. ASML’s moat and constrained supply argue its upside is underappreciated vs headline-driven AMZN selloffs; historical parallel: early cloud capex cycles where equipment vendors outperformed hyperscalers post-capex surge. Unintended consequence: aggressive cost cuts at healthy firms may boost short-term margins but permanently erode growth engines—favor concentrated asymmetric hedges rather than large outright shorts.

Key Decisions for Investors

• Establish a 2–3% portfolio long position in ASML (ASML) within 7–14 days; add another 1–2% on a >7% pullback or after a quarterly order-book beat; set a protective stop at 12% below entry and target +30% over 12–18 months.
• Initiate a 1.5–2% portfolio position via a 3–6 month AMZN (AMZN) put spread (buy nearer‑ATM put, sell lower strike to cap cost) to express downside from capex/FCF risk; exit if AWS revenue growth beats consensus by >200bps or management cuts 6–12 month capex guidance by >10%.
• Implement a pair trade: long ASML vs short AMZN equal notional (2% long ASML financed by 2% short AMZN) to capture relative outperformance; unwind if the ASML:AMZN price ratio diverges >20% from entry or on confirmed orderbook deterioration (>20% q/q decline).
• Rotate 3–5% of portfolio from consumer discretionary and unprofitable growth names into semiconductor capital-equipment suppliers and copper/mining names over 30–90 days; monitor Dutch/US export-policy updates (trigger: any new restrictions within 60 days) and next ASML/AMZN earnings for rebalancing.