New financial data reveals a notable deceleration in business spending on artificial intelligence, even as the volume of AI usage continues to expand. Ramp, a financial technology company, reported that AI token spending across its platform saw only a 1.5% increase in August and September 2026. This follows a substantial 42% month-over-month growth in July, indicating a sharp shift in spending patterns. The current trend challenges the prevailing assumption that increased AI usage directly translates into equivalent revenue growth for AI model providers.
The slowdown in spending is attributed to a combination of falling prices for AI services and more rigorous cost management by large corporate clients. Between late September and late October 2026, the average price per million tokens for services from OpenAI and Anthropic decreased by 26.8%. OpenAI saw a 42.4% price drop, while Anthropic's prices fell by 15.7%. These price reductions help explain why a significant rise in AI token volume can result in a much smaller increase in overall expenditure.
Furthermore, a concentrated group of major AI spenders disproportionately influences the broader market trends. Ramp's analysis from October 14, 2026, indicated that 1% of firms accounted for 80% of spending on OpenAI and Anthropic, and these firms experienced nearly flat token expenditure for two consecutive months. In September 2026, the median monthly AI spending per employee within this top 1% of firms was $9,568, starkly contrasting with $15.70 for the median firm across Ramp's network. This concentration means that the cost-saving measures implemented by a few large customers can significantly temper overall spending growth, even as AI adoption broadens.
Despite the spending slowdown, AI adoption continues to widen. In September 2026, 55.9% of businesses on Ramp were paying for AI services, a figure that has grown steadily. Anthropic has captured a larger share of this spending, accounting for 60.5% of AI token expenditures in the week ending October 23, 2026, compared to OpenAI's 35.1%. This shift in market share demonstrates that a provider can increase its slice of a shrinking or slower-growing spending pie through efficiency gains or competitive pricing.
This evolving financial landscape for AI companies comes as Anthropic, a prominent artificial intelligence firm, prepares for its initial public offering. The company publicly filed its Form S-1 registration statement with the U.S. Securities and Exchange Commission on October 26, 2026, indicating its intention to list on the Nasdaq Global Select Market under the proposed ticker symbol "ANT." The filing disclosed significant financial commitments, including approximately $321.4 billion in leases that had not yet commenced as of June 30, 2026, with terms extending up to 20 years. While Anthropic reported $16.3 billion in revenue for the first half of 2026, it also recorded a net loss of $80.7 billion for the same period, largely influenced by changes in the fair value of convertible bonds. The company's substantial long-term lease obligations and capital expenditures highlight the significant infrastructure investments required to scale AI operations. The market will be closely watching how investors perceive these financial commitments and growth projections in light of the current trend of slowing AI spending.
