Menlo Ventures published its 2026 State of Enterprise AI report on November 3, 2026. The data, based on a survey of 400 US decision-makers, reveals that corporate AI spending significantly exceeded initial forecasts. While companies planned for a 1.9-2x increase, actual expenditure reached 2.5x of the budgeted amounts.
Total enterprise AI deployment spending hit $176.8 billion by September 2026, compared to $33 billion in 2025. This represents a 5.4x increase. A small group of heavy users running autonomous AI agents drove this growth. These agents now account for 80% of token usage on model interaction platforms, with activity increasing 37 times since early 2026.
The report highlights that 55% of the increased spending originated from new budgets, while only 9% was reallocated from traditional SaaS tools. The share of enterprises investing in AI grew from 55% to 65%. Infrastructure spending was concentrated among three major labs: Anthropic, OpenAI, and Google, which captured 69% of the total deployment spend, or $122.7 billion.
Within the large language model API market, Anthropic leads with a 55% share, followed by OpenAI at 31% and Google at 9%. Spending on AI applications reached approximately $77 billion, up from $19 billion in 2025. Coding tools accounted for $34 billion of this, while productivity tools totaled $25 billion, together representing 76% of application-related expenditures.
