The $485-billion Question

The evidence is accumulating, not arriving all at once. That is the central tension running through Goldman Sachs' August 2026 AI Adoption Tracker¹, published September 1, 2026 by economists Sarah Dong and Joseph Briggs. The picture they construct is of an AI buildout that remains one of the most consequential capital cycles of the modern era, paired with an adoption curve that, while rising, has yet to f`ully translate that investment into economy-wide productivity gains. The divergence between those two trajectories is precisely where the most important near-term signals for advisors and investors reside.

The Scale of the Buildout

Dong and Briggs state plainly that "AI-related investment growth remains strong, particularly for semiconductors," where global revenue is projected to reach $863 billion annualized by year-end. AI hardware investment in the US national accounts now stands $485 billion above 2022 levels on a three-month annualized basis, roughly 1.5% of GDP. Gross profit forecast revisions for AI-exposed global public companies since 2022Q3 have accumulated to $1.1 trillion. Taiwan's AI-related manufacturing shipments have grown 76% since 2022. US AI hardware net imports reached $55.9 billion in June alone. These are not speculative commitments. They are capital already deployed and infrastructure already built.

The Adoption Picture

AI adoption among US establishments rose 0.9 percentage points to 22.4% in August 2026, according to the Census Bureau's Business Trends and Outlook Survey, with firms expecting that figure to reach 25.9% within six months. The Information sector is approaching 50% adoption. Finance and Insurance posted the largest single-period gain in the latest update. At the subsector level, computing, data hosting, and publishing continue to lead. Warehousing and miscellaneous manufacturing report the largest expected increases ahead, suggesting the wave is moving down the value chain into more operationally intensive sectors.

Size matters considerably. Establishments with more than 250 employees lead adoption at roughly 40%. Smaller firms have lagged meaningfully. That divergence implies the productivity dividend will concentrate first in larger, better-resourced organizations before diffusing broadly. Separately, 40% of large surveyed organizations report actively deploying and scaling AI agents, a qualitative shift from AI as a discrete tool toward AI as an embedded participant in operational workflows.

Labor: Visible, But Not Yet Systemic

Dong and Briggs are deliberate in their framing here. The labor market impact, they write, "remains visible but narrow." Employment drags are observable in marketing, graphic design, customer service, and parts of the tech sector. Those losses are being partially offset by construction job growth in data center-exposed categories, up 209,000 since 2022 relative to broader construction trends. Corporate layoffs explicitly attributed to AI reached 11,000 in June, bringing the 2026 year-to-date total to 113,000. The tech sector's share of total employment continues to decline relative to its pre-2022 trend. The correlation between AI adoption and broad labor market slack has strengthened in recent months, though it has not yet reached statistical significance. Twenty-two percent of Russell 3000 companies mentioned AI and labor-related keywords in 2026Q2 earnings calls, a figure that has risen sharply from near zero just a few years ago.

Where Productivity Lives

The productivity evidence, for now, remains concentrated in the areas of deepest deployment. Dong and Briggs observe "large impacts on labor productivity in the limited areas where generative AI has been deployed but only modest signs of economy-wide impacts." Academic studies point to an average productivity uplift of 23%. Company anecdotes suggest efficiency gains closer to 32%. Critically, industries with higher adoption rates are now showing a slight but measurable acceleration in productivity growth in official US data. That is the signal worth tracking closely.

Five Key Takeaways for Advisors and Investors

1. AI investment is structural, not cyclical. Hardware investment running at 1.5% of GDP above 2022 levels, combined with $1.1 trillion in gross profit forecast revisions for AI-exposed public companies, signals a capital cycle with durable momentum. Semiconductor exposure remains central to that story.
2. Adoption is rising but unevenly distributed. Large firms lead; smaller firms lag. The competitive divergence this creates across industries is a consideration worth building into portfolio and business analysis.
3. Labor market impact is real but contained, for now. Specific occupations face meaningful displacement. The aggregate signal is strengthening but has not tipped into broad disruption. The trajectory warrants ongoing attention as adoption rates continue climbing.
4. Productivity gains are emerging at the industry level. Sectors where AI deployment is deepest are beginning to register measurably better outcomes in official productivity data. If sustained, that signal carries significant implications for long-run earnings growth.
5. AI agent deployment marks a qualitative inflection. Forty percent of large organizations are deploying and scaling agents. This transition, from AI as assistant to AI as workflow participant, changes how firms should think about technology investment, human capital strategy, and competitive positioning over the coming years.

Footnote:

1 Dong, Sarah, and Joseph Briggs. "AI Adoption Tracker: August 2026: Adoption Rises to 22.4%." Goldman Sachs Global Investment Research, 1 Sept. 2026, https://www.gspublishing.com/content/research/en/reports/2026/09/01/d55f74c5-6e25-4bc3-98e4-dac4da447611.pdf.