AI Now Powers 40% of U.S. GDP Growth, NYC Report Claims

TL;DR

• New York City Comptroller’s Office reports AI spending on equipment, software, and R&D accounted for nearly 40% of U.S. real GDP growth in the first three quarters of 2025.
• The report models a 10-year outlook for how AI adoption will reshape NYC’s tax base, labor market, and sector mix — warning of both productivity gains and major job displacement.
• Economists push back, arguing that crediting a single technology for nearly half of national growth is methodologically shaky and risks obscuring structural inequality.
• The findings fuel Wall Street’s narrative that AI capex is the engine behind the current equity boom, justifying massive chip and cloud spending by NVIDIA, Microsoft, Google, and AWS.

New York City Puts a Number on AI’s Economic Punch

The New York City Comptroller’s Office just dropped one of the first official U.S. government analyses quantifying how much of the recent economic expansion runs on AI. According to the report, nearly 40 percent of real GDP growth during the first three quarters of 2025 was directly attributable to AI investment in equipment, software, and R&D. That’s not a tech industry trade group estimate or a venture capital pitch deck. That’s a city government trying to figure out what happens to its budget when the economy’s center of gravity shifts.

The report doesn’t stop at the macro picture. It models a 10-year time horizon to project how AI adoption will reshape New York City’s fiscal outlook — its tax revenues, labor market composition, and the high-wage sectors like finance, media, and professional services that anchor the city’s economic base. The message is clear: AI isn’t just a productivity story anymore. It’s a municipal budget story, a tax policy story, and a labor displacement story all at once.

New York isn’t guessing. It’s planning for a world where AI capex is the single biggest driver of growth, and where the fallout — both upside and downside — lands hardest in cities that depend on knowledge work.

Why NYC’s Comptroller Is Sounding the Alarm on Job Reshuffling

Here’s what makes this report different: it’s not written by economists trying to model the national economy in the abstract. It’s written by a city comptroller trying to figure out whether tax revenues will cover next decade’s payroll. And the answer depends entirely on whether AI creates more high-wage jobs than it destroys — and whether those jobs stay in New York.

The report flags productivity gains as a double-edged sword. Sure, AI-driven efficiency could boost output per worker and lift wages in sectors that adopt it early. But it also warns of significant job reshuffling — a polite way of saying that entire categories of white-collar work could vanish or compress into lower-paid roles. Finance, media, legal services, consulting — these are the sectors that keep New York’s tax base solvent. If AI automates the associate attorney, the junior analyst, and the copywriter, the city’s revenue model breaks.

I’ve covered enough automation waves to know that ‘reshuffling’ is the word policymakers use when they don’t want to say ‘displacement.’ And the comptroller isn’t sugarcoating it. The report treats AI as both an engine of growth and a wrecking ball for labor markets that haven’t diversified beyond high-cost knowledge work.

Think of it like this: AI investment is a flood. It lifts all boats — until you realize half the boats are anchored to docks that are underwater. New York’s economy is one of those docks. The city bet big on industries where human expertise commanded a premium. AI doesn’t care about premiums.

But let’s zoom out. If AI really did drive nearly 40% of GDP growth in the first three quarters of 2025, that’s not just a New York problem. That’s a national economic story that hasn’t fully registered outside of Silicon Valley and Wall Street. We’re talking about a technology that went from experimental to macroeconomically significant in less than three years. No other general-purpose technology — not the internet, not mobile — hit that velocity.

And yet economists are skeptical. They argue that attributing growth to a single technology is methodologically tricky. They’re right. GDP accounting is messy. Separating AI investment from broader software capex, cloud infrastructure spending, and R&D that would have happened anyway is hard. Over-indexing on AI could obscure other structural factors — like post-pandemic supply chain normalization or fiscal stimulus effects — that also juiced growth. Worse, it could lead policymakers to underestimate distributional impacts on workers who don’t benefit from productivity gains.

Fair criticism. But here’s the counterargument: if the comptroller’s math is even half right, we’re already in a world where AI capex is the dominant marginal driver of growth. Quibbling over methodology misses the point. The policy question isn’t whether AI accounts for 40% or 30% of GDP growth. It’s whether cities and states are prepared for an economy where capital investment in AI crowds out labor income — and what that means for tax bases built on payroll and property.

How Wall Street and Silicon Valley Are Reading the Data

This report lands in the middle of a debate that’s been raging on Wall Street for months: is the AI capex boom sustainable, or is it a bubble waiting to pop? The comptroller’s findings feed directly into the bull case. If AI investment is already driving 40% of GDP growth, then the current equity rally — powered by NVIDIA’s chip sales, Microsoft’s Azure AI revenue, Google’s Gemini bets, and AWS’s infrastructure build-out — isn’t speculative. It’s rational.

NVIDIA’s stock has reportedly tripled since early 2024, and the company’s data center revenue has become the single largest component of U.S. semiconductor exports. Microsoft and Google are each spending tens of billions annually on AI infrastructure. AWS is racing to match them. The comptroller’s data suggests that spending isn’t getting ahead of demand — it’s meeting it. And if nearly half of recent GDP growth traces back to AI, then the firms selling the picks and shovels are in the right place.

But that narrative has a dark side. If AI investment is the only thing driving growth, what happens when the capex cycle peaks? Or when productivity gains don’t translate into wage growth because automation compresses labor demand? The comptroller’s report doesn’t answer those questions. It just makes them more urgent.

Competitive context matters here. The U.S. isn’t the only country pouring capital into AI. China’s government is bankrolling its own AI infrastructure push, and European cloud providers are trying to carve out regional sovereignty. If AI capex is the engine of GDP growth, then the countries that control the supply chain — chips, models, cloud capacity — control the growth itself. That’s a geopolitical story as much as an economic one.

What This Means for Cities That Bet on Knowledge Work

New York City isn’t unique. San Francisco, Seattle, Boston, Austin — every city that built its tax base on high-wage knowledge work is now staring at the same question: what happens when AI makes that work cheaper, faster, or unnecessary? The comptroller’s report is a roadmap for how local governments should think about AI, and it’s not optimistic by default.

Following the pandemic, U.S. productivity growth had been sluggish. Economists worried that the post-2008 productivity slowdown was permanent — that we’d entered an era of secular stagnation where growth would stay low and labor markets would stay tight without wage gains. Then AI capex surged. Suddenly productivity is back, but it’s concentrated in firms that can afford billion-dollar model training runs and million-dollar inference infrastructure. That’s not broad-based growth. That’s winner-take-all.

New York’s comptroller is trying to game out what that means for a city where finance, media, and professional services are the economic base. If AI automates the junior banker, the paralegal, and the marketing analyst, does the city lose tax revenue? Or does it gain revenue from higher corporate profits and more efficient firms? The answer depends on whether productivity gains get shared with workers or captured by shareholders.

And here’s the kicker: the report uses a 10-year time horizon. That’s not a long-term forecast. That’s a medium-term budget planning window. The comptroller isn’t worried about what AI does in 2050. The comptroller is worried about what AI does by 2035 — and whether the city’s fiscal model survives it.

Three Things to Watch as Cities Model AI’s Fiscal Impact

First, watch whether other major cities follow New York’s lead and publish their own AI fiscal impact reports. If Los Angeles, Chicago, and San Francisco start modeling how AI reshapes their tax bases, that signals a shift from treating AI as a tech story to treating it as a core municipal finance issue. Cities that wait too long to plan will get caught flat-footed when revenue models break.

Second, track whether the comptroller’s 40% figure holds up as more quarterly GDP data rolls in. If AI investment’s share of growth stays near 40% through the end of 2025 and into 2026, that cements AI as the dominant driver of the current expansion. If it drops sharply, it suggests the capex boom is peaking — and cities betting on sustained AI-driven growth need to revise their models.

Third, monitor how federal and state policymakers respond to the labor displacement warnings buried in this report. If AI is driving growth but also triggering significant job reshuffling, that creates pressure for new safety nets, retraining programs, or tax reforms that capture more revenue from AI-driven productivity. The comptroller’s report is a fiscal document, but it’s also a political signal. Cities are starting to ask whether AI’s winners should pay more to support the workers AI displaces.

Source: New York City Comptroller