US recession risk is moving back to center stage as the economy’s surprising resilience shows early cracks and the AI surge that lifted growth begins to cool. After years of defying ominous signals—an inverted yield curve, soft leading indicators and shaky sentiment—markets are confronting a tougher setup built on slower hiring, rising delinquencies and questions about whether massive AI spending can keep delivering.
The headline picture still looks solid. Unemployment remains low by historical standards, major indexes hover near highs and growth has held up better than many expected.
Beneath the surface, the tone is changing. From corporate capex discipline to consumer strain, the list of caution flags is getting longer.
Why US recession risk is rising despite recent strength
For much of the past two years, the economy outran a classic warning package: an inverted yield curve, weakening leading indicators and deteriorating business and consumer confidence. That unusual gap between signals and outcomes was powered in part by a historic AI investment cycle and stronger than expected household balance sheets.
Today, US recession risk appears to be climbing for several reasons:
- Hiring has downshifted. Recent monthly payroll gains have slowed sharply from last year’s pace, reflecting softer demand for labor across several sectors.
- Credit stress is building. Delinquencies on credit cards and auto loans past 90 days have climbed toward levels last seen during the financial crisis, pointing to household strain.
- Industrial activity is weak. The ISM Manufacturing Index has spent much of the year below 50, a contraction signal, while exports and production face tariff headwinds and softer global demand.
- Inflation is sticky in key categories. Groceries, housing and health care remain expensive even as headline inflation cools, squeezing real purchasing power.
- Policy uncertainty is high. Tariff announcements and a government shutdown have complicated planning for companies and consumers.
Historically, the inverted yield curve has flagged US recession risk well ahead of an actual downturn. The fact that expansion persisted despite such warnings has been unusual. With the AI impulse maturing and consumers feeling the pinch, the gap between signals and growth may narrow.
The AI boom that powered growth
The AI cycle, ignited after late-2022’s breakout of ChatGPT, changed the investment landscape almost overnight. Companies across industries rushed to experiment with generative AI to avoid falling behind, triggering a global arms race in model training, data infrastructure and specialized chips.
- According to Stanford University’s 2025 AI Index Report, AI investments in the US surged to $109.1 billion in 2024 alone.
- For perspective, total AI investments from 2013 to 2023 were $361.8 billion, or about $33 billion per year on average.
That wave of spending did heavy lifting. It helped offset softness in manufacturing, housing and commercial real estate, keeping GDP on a firm footing even as traditional cyclical engines sputtered.
Public markets followed the money. AI enthusiasm propelled a handful of mega-cap names. Nvidia rallied more than twelvefold since late 2022. Microsoft, Alphabet, Amazon and Meta doubled or more as investors priced in years of AI-driven growth.
The concentration effect was powerful. A narrow group carried the Nasdaq and S&P 500 higher, even as many non-tech stocks lagged.
Can tech’s capital spending keep paying off?
As the cycle matures, the conversation is shifting from buildout to payback. The question for boards and investors is simple: Do the returns justify the unprecedented spend.
- A study by MIT found that roughly 95 percent of organizations reported no measurable return on AI investments to date.
- A McKinsey survey indicated most firms saw cost savings under 10 percent, with revenue gains typically below 5 percent.
At the same time, big tech leaders have signaled a willingness to lean in. Meta’s Mark Zuckerberg said he would rather “misspend a couple hundred billion dollars” on AI than risk falling behind competitors, underscoring the strategic urgency driving capex.
If productivity gains remain hard to pin down, companies may slow the pace of investment to align with realized returns. That would remove a major tailwind that masked weakness elsewhere and could raise US recession risk if AI spending normalizes faster than end-user demand materializes.
Cracks beyond tech: households, jobs and industry
Outside of AI, more traditional gauges are cooling.
- Job growth has softened, with nonfarm payroll gains averaging roughly 75,000 per month from January through August, down from 168,000 last year.
- The Federal Reserve cut rates in September to cushion the slowdown, even as some economists warn new tariffs could rekindle price pressures.
- Household stress is visible in rising delinquencies and thinner savings buffers, particularly for lower and middle-income consumers.
Manufacturing remains under pressure. The ISM index has lingered below 50 for most of the year, while exports and industrial production face weaker global demand and higher costs tied to trade policy. Business investment outside AI looks sluggish as companies delay projects in the face of uncertain policy and a cloudier outlook.
If these trends persist, US recession risk will remain elevated as multiple engines of growth run below potential.
Credit stress and corporate defaults back on the radar
Stress in corporate credit markets is another area to watch. Defaults have begun to climb from unusually low levels, with recent bankruptcies including subprime auto lender and dealer Tricolor and auto parts supplier First Brands.
JPMorgan CEO Jamie Dimon sounded a familiar caution: “When you see one cockroach, there are probably more.” Investors read that as a reminder that credit problems often surface unevenly before they broaden.
Historically, broader credit deterioration can tighten financial conditions and amplify US recession risk if lending standards harden and investors demand higher compensation for risk.
Are circular AI funding loops masking real demand?
Another emerging worry is whether parts of the AI economy are sustained by circular funding arrangements rather than durable customer demand.
- Reports in September suggested Nvidia could invest up to $100 billion in OpenAI’s data center buildout, with OpenAI then purchasing large volumes of Nvidia chips for those facilities.
- Subsequently, OpenAI agreed to spend tens of billions on AMD chips in a deal that also gives it the option to acquire a meaningful stake in AMD at a nominal price.
- OpenAI has also struck multibillion-dollar infrastructure agreements with Oracle and CoreWeave, both scaling capacity with Nvidia GPUs.
These intertwined deals can make it difficult to separate true end-user demand from financing loops. Bloomberg has reported that OpenAI remains unprofitable, with revenue estimated to rise from $3.7 billion in 2024 to about $13 billion in 2025 while losses expand from roughly $5 billion to around $8 billion. That reporting suggested the company may not turn cash-flow positive until later in the decade if projections hold.
If capital markets become less forgiving or returns stay muted, a pullback in AI capex could ripple across chips, cloud, software and data center real estate. That would raise US recession risk if the AI leg of the expansion wobbles at the same time household and industrial demand fade.
Growth outlook: slower from here
Forecasts compiled by private-sector economists and the research cited here point to slower growth ahead. US GDP is projected to hover near 1.8 percent this year and next, down from an estimated 2.8 percent in 2024 as AI tailwinds moderate, hiring cools and policy uncertainty curbs investment.
None of that guarantees a downturn. A modest expansion is still possible if productivity improves and inflation continues to ease. But the balance of risks has shifted, and markets are more sensitive to negative surprises than they were during the height of the AI surge.
US recession risk tends to climb when multiple pillars soften together—jobs, credit, capex, exports—and when policy clarity is lacking.
What a slowdown could mean for markets
Equities have proven resilient, but the setup is more fragile.
- Valuations remain elevated for many AI leaders relative to historical norms, leaving less margin for disappointment.
- If AI capex moderation arrives before revenue catch-up, multiples could compress in semis, cloud infrastructure and select software names.
- Defensive sectors can provide ballast, though they are not immune if earnings expectations reset.
For fixed income, slower growth with easing inflation supports duration as long as inflation shocks do not return. Credit selection matters more as defaults rise from trough levels.
Currency and commodities will take their cues from the growth-inflation mix and policy paths. Tariff risk adds another layer for trade-exposed sectors.
In short, market leadership may broaden if AI favorites consolidate, or narrow further if investors keep paying up for perceived defensibility. Either way, US recession risk is a key variable in how that leadership evolves.
Scenarios investors are modeling
- Soft-landing glide path: Inflation cools further, productivity lifts from targeted AI deployments and growth stabilizes near 2 percent. US recession risk recedes into 2026.
- Shallow recession: Hiring slows further, AI capex resets and credit stress broadens. Earnings dip modestly, the Fed eases more, and recovery begins within a few quarters.
- Stop-start cycle: Inflation flares episodically due to tariffs or energy shocks, complicating policy and keeping volatility elevated as growth hugs the zero line.
Positioning across these paths hinges on how quickly AI spend converts to measurable returns and whether consumer strain eases.
Energy costs and the AI footprint
Electricity is emerging as a pressure point for households and industry. The buildout of energy-intensive data centers for AI is set against decisions that raised costs for some power equipment and delayed certain renewable projects. Households already report higher utility bills in many regions.
Relief will likely depend on grid upgrades, diversified generation and efficiency gains. If power prices keep rising, companies could revisit energy-intensive projects and households could pull back on discretionary spending, nudging US recession risk higher.
Policy watch: trade, rates and data
Policy choices will shape the next leg.
- Trade tariffs and exemptions: Clarity on scope and timelines can reduce business uncertainty. Persistent ambiguity dampens investment.
- Federal Reserve: The September rate cut was a signal that policymakers are leaning against slowdown risk. The pace from here depends on inflation and job trends.
- Government operations: Shutdowns that delay critical data releases leave markets and the Fed flying blind. Timely data is essential for good decisions.
A more predictable policy path would help businesses plan capex and hiring, while reducing the chance that uncertainty itself amplifies US recession risk.
What to watch next: data, earnings and credit
Investors tracking the cycle are focused on a few high-frequency gauges:
- Labor: Payroll growth, jobless claims, labor-force participation and wage trends
- Prices: Core services inflation, shelter disinflation and market-based inflation expectations
- Manufacturing: ISM new orders, export demand and inventories-to-sales ratios
- Credit: High-yield spreads, bank lending standards, small business loan demand and default rates
- AI indicators: Cloud capex plans, chip order visibility, data center utilization and early productivity wins at the customer level
Evidence that AI deployments are delivering tangible productivity gains would reduce US recession risk by supporting margins and real wages. Signs of capex pullback without offsetting demand would point the other way.
Practical considerations for readers
While every portfolio and household is different, a few principles are widely applicable in a late-cycle environment:
- Revisit diversification. Avoid overconcentration in any single theme, even one as powerful as AI.
- Emphasize quality. Balance sheets, cash flow visibility and pricing power matter more when growth slows.
- Stress-test budgets. Account for higher utilities and the possibility of slower income growth.
- Keep liquidity. Emergency funds and portfolio liquidity dampen the need to sell at unfavorable times.
These steps do not predict outcomes. They improve resilience across a range of plausible scenarios that include elevated US recession risk.
Bottom line
The economy and markets outran a formidable set of warning signals thanks to an extraordinary AI cycle and lingering post-pandemic tailwinds. That phase may be passing. Hiring is cooler, credit is wobbling and the payback on large AI investments is under sharper scrutiny.
If AI-driven productivity gains become more visible and price pressures continue to fade, the expansion can continue. If capital spending resets before demand catches up and household stress intensifies, US recession risk will rise and market leadership could shift.
Either way, the next stretch looks more nuanced than the last. Staying grounded in data, watching the AI return-on-investment story and maintaining flexibility are key as the cycle evolves.
Article Source: inquirer.net
