The American economy is growing, but the footing looks a lot less comfortable.
GDP expanded at a 1.5% annualized pace in Q2, down from 2.1% in Q1, while July payrolls dropped by 23,000 and unemployment held near 4.1%, as reported by CNBC.
Yet private domestic demand jumped to 4.2%, as reported by Investing.com, suggesting that the engine hasn’t stalled. Goldman Sachs CEO David Solomon argues a bigger story is developing beneath these mixed signals.
Consumers are at the heart of that dichotomy.
Personal income increased by 0.4% in July, but real spending was mostly flat, the saving rate tanked to 3%, and retail sales fell 0.6%. Inflation also remains sticky, with the PCE index running at 3.7% above last year, as reported by Reuters.
Then we have the enormous AI investment cycle. Companies continue to borrow heavily for funding infrastructure, raising questions about whether the boom might eventually create a credit problem or lead to a painful reset.
Solomon’s remains unexpectedly constructive.
In a CNBC interview, he talked about how he sees enough strength beneath the surface to look past the current risks.
His more striking argument is that a familiar technology might unleash an “extraordinary” productivity boom that could potentially revamp America’s long-term growth potential.
Why Solomon sees a stronger U.S. economy beneath the noise
Solomon argues that even though the current headwinds might slow down the economy, he doesn’t believe they have broken its engine.
That distinction effectively shapes how he interprets every major risk ahead.
More Economy:
Kevin O’Leary raises stark concern about inflation
Goldman Sachs delivers its verdict on inflation and jobs
Bank of America issues stark warning on Fed and economy
“Generally speaking, you know, the consumer is still pretty resilient,” Solomon said. “The economy is performing well.”
It’s important to note that consumer spending remains the biggest source of U.S. activity. If households continue spending while labor conditions cool off, the economy could absorb weaker patches without tipping into a contraction.
The second pillar is the investment cycle.
Solomon pointed to an “enormous investment cycle” contributing to growth and activity, a nod to the heavy spending surrounding AI, data centers, and related infrastructure. According to him, the capital buildout is supporting demand today, even before the promised productivity gains arrive.
Corporate earnings reinforce the argument.
Solomon called the numbers coming in as “extraordinary,” describing it as “a big tailwind for the market and for the economy.” Rising profits offer businesses greater capacity to invest, hire, service debt, and absorb higher borrowing costs.
The broader data support that argument.
U.S. corporate profits from current production shot up by $400.9 billion in Q2, which is more than five times the $74.4 billion increase recorded in Q1.
Moreover, FactSet’s latest analysis showed that the Magnificent 7 stocks delivered 118.5% earnings growth in Q2, while the other 493 S&P 500 companies posted blended growth of 31.8%, their strongest pace since late 2021.
Nevertheless, Solomon didn’t predict a frictionless outlook, either.
He acknowledged “bumps and headwinds” from the Middle East, arguing that the trade policy and tariffs remain “a headwind to some degree.” Nevertheless, he sees these as obstacles that the economy can efficiently work through.
Perhaps the long-term case is a lot more ambitious.
As AI moves into enterprises, Solomon forecasts an “extraordinary” productivity boom that could potentially create “a fundamentally higher growth rate.”
Goldman Sachs CEO David Solomon believes the current headwinds impacting the economy haven’t broken its engine.Brendon Thorne/Bloomberg via Getty Images
AI productivity boom is showing up, but only in pieces
Solomon’s AI optimism isn’t entirely built on promise.
We’re seeing some early evidence that the technology is saving time, elevating less-experienced workers and spreading quickly. The bigger question, though, is whether those isolated gains could become an economy-wide acceleration.
The bull case starts with the numbers.
U.S. nonfarm business productivity rose 2.2% year over year in Q2 2026. Since late 2019, it has risen at a 2.1% annualized rate, above the 1.5% pace of the previous business cycle. Though it’s unfair to attribute the lion’s share of that improvement to AI, the timing is consistent with an emerging contribution.
Moreover, workplace evidence is a lot more direct.
St. Louis Fed data showed that 39.2% of employed adults used generative AI at work by Q2, up substantially from 28.2% in Q3 2024. AI-assisted hours increased to 6.3% of total work time, while reported time saved reached 2.2%.
Separately, an NBER field study revealed that AI raised customer-support productivity nearly 14%, with the biggest gains among less-experienced workers.
That said, task-level efficiency still hasn’t produced a visible macroeconomic boom. Productivity rose only 1.4% annualized in Q2, behind its long-run 2.1% pace. Manufacturing productivity grew just 0.5% annually during the current cycle, underscoring that the gains remain concentrated in cognitive services.
Diffusion remains another constraint.
Census data show that just 17% to 20% of U.S. businesses used AI through early May 2026. A Danish study covering 25,000 workers found chatbots saved nearly 3% of time but had no significant effect on earnings or recorded hours.
On top of that, MIT economist Daron Acemoglu estimates that AI might raise total factor productivity by no more than just 0.66% over a decade, or even less than 0.53% if more difficult tasks are considered.
What Solomon’s outlook means for investors
Solomon’s comments offer investors more reason to stay constructive, but it’s wise not to ignore valuation or execution.
The pillars he talks about can extend the cycle, but a lot of that optimism is priced into years of AI-powered growth. That said, it’s imperative to distinguish between the spending beneficiaries and productivity beneficiaries.
Chipmakers, data-center operators, and power suppliers are monetizing the buildout now. However, it remains critical for their customers to prove that their AI spending will translate into lower costs, higher growth, or wider margins. If those returns emerge, AI could effectively broaden out earnings beyond infrastructure leaders.
Balance sheets also matter. Solomon feels that there’s little systemic credit risk because many of the largest borrowers generate a ton of cash flow.
A useful example is Google parent Alphabet (GOOG) stock. Despite a massive $44.9 billion capex in Q2, primarily for AI infrastructure, it generated $185.7 billion in trailing 12-month operating cash flow. It also maintained $242.5 billion in cash and securities, along with $98.2 billion in long-term debt.
That could reduce near-term danger, but it doesn’t eliminate misallocation. Investors need to continue monitoring free cash flow after considering capital spending, debt growth, interest coverage, and whether AI sales can scale more quickly than depreciation and financing costs.
Perhaps the macro signal is productivity. If we see sustained output-per-hour growth over the pre-pandemic trend, it would support higher profits without reigniting inflation, potentially paving the way for strong growth alongside easier monetary policy.
The obvious conclusion is selective optimism. Favor companies already converting AI into measurable sales, margins or customer savings, while treating distant productivity promises cautiously.
Related: Scott Bessent has surprising answer for U.S. debt fears