Corporate America’s AI Spending Gap: Top 1% Invest $7,400 Per Employee

The top 1% of U.S. businesses are investing a median of $7,400 per employee in artificial intelligence as of July 2026, creating a stark divergence in operational spending. According to the Ramp AI Index, this elite tier outspends the median company—which allocates just $11.95 per employee—by a factor of 600, signaling a deepening divide in corporate resource allocation.

The Growing Divide in AI Capital Expenditure

Corporate investment in generative AI is currently defined by two distinct speeds. While the top 1% of firms commit significant capital, the top 10% of businesses spend roughly $650 per employee. The Ramp AI Index, released August 12, 2026, highlights that even among aggressive adopters, the spending gap is substantial—with the top 1% outspending the next tier by roughly 11 times.

Anthropic Leads in Market Share as Model Pricing Hits a Ceiling

Anthropic currently holds the largest share of the business AI market, with 43.5% of U.S. businesses subscribing to its models as of July, an increase of 1.1 percentage points from the previous month. OpenAI, by contrast, grew by only 0.23 percentage points, while xAI reached 4% of the market.

Priced at approximately $10 per million tokens—double the rate of OpenAI’s GPT-5.6 Sol—Fable 5 struggled to capture significant volume. One month post-launch, Fable 5 accounted for only 6% of tokens purchased from Anthropic and 11.4% of total dollar spend. In comparison, GPT-5.6 Sol captured 25% of OpenAI’s tokens and 23% of its spend. “With Fable 5, we’ve found a new upper bound for how much businesses are willing to spend on AI,” notes Ara Kharazian, lead economist at Ramp. This sensitivity to pricing is driving a slow but steady shift toward open-source and Chinese-developed models, which saw adoption rise to 6.1% in July.

Organizational Readiness as the Primary Performance Constraint

The financial payoff of these investments remains difficult to quantify. According to the PYMNTS Intelligence Enterprise AI Benchmark Report, 71% of senior technology executives at companies with at least $1 billion in annual revenue identify organizational readiness—not the technology itself—as the primary barrier to performance. Only 11% of these executives cite the underlying AI models as a limiting factor.

A Goldman Sachs analysis cited by PYMNTS indicates that only 2% of S&P 500 companies quantified the effects of AI in their second-quarter earnings reports. Among those that did, 11% reported measurable productivity gains. Interestingly, the companies that successfully quantified these metrics saw median earnings rise by 17%, compared to 14% for those that did not.

Shifting Expectations for Return on Investment

CFO sentiment toward AI payoff timelines has shifted dramatically since mid-2025. Data from PYMNTS Intelligence shows that the share of CFOs expecting positive returns within one to two years has surged to 39.1%, up from effectively zero. Simultaneously, those anticipating a three-to-five-year horizon have dropped from 65.9% to 34.8%.

As the market enters this maturation phase, the focus will likely shift from broad-spectrum adoption to targeted, high-margin use cases.