Key takeaways
- Microsoft and Meta are reportedly cutting AI tool spending
- Enterprise AI spending growth is slowing, with new Ramp data showing that businesses are becoming more cautious about costs
- The AI boom is facing financial pressure as infrastructure delays, rising costs and uncertain returns push companies and investors to prioritize profitability
Companies are beginning to rein in artificial intelligence spending as soaring costs, disappointing returns and growing concerns about the sustainability of the AI boom are pushing businesses to reassess their investments.
Microsoft and Meta are reportedly reducing spending on third-party AI tools, while new research suggests enterprise AI expenditure is losing momentum despite continued adoption.
The shift is happening as investors question whether billions of dollars poured into AI development and infrastructure will actually generate good enough returns.
Microsoft and Meta Reportedly Slash AI Tool Spending
Microsoft reportedly reduced internal spending on Anthropic's Claude by more than one-third from a projected annual bill exceeding $1 billion. This is according to reports published in early October.
The technology giant is instead encouraging employees to use GitHub Copilot.
Meanwhile, Meta reportedly halved its Claude Code users from approximately 60,000 to 30,000, favoring internally developed alternatives, including MetaCode and Muse Code.
This suggests that even companies leading the AI revolution are becoming more selective about which services justify their costs. However, these reductions primarily concern third-party software rather than overall AI investment.
New Research Shows AI Spending Growth Is Slowing
The latest Ramp AI Index that was published Oct. 14 provides even more evidence of slowing enterprise expenditure. The report found that AI adoption among businesses using Ramp reached 55.9% in September, although growth in new customers slowed considerably.
More significantly, spending per employee among the top 1% of AI-spending companies stood at $9,568 in September, approximately 3.1% below July's peak.
(Source: ramp.com)
Businesses also recorded two consecutive months of virtually no growth in AI token spending, despite increasing usage.
Ramp attributed the slowdown partly to falling model prices and improving efficiency. This allows companies to accomplish more without increasing their budgets.
AI Infrastructure Boom Faces Growing Financial Pressure
Concerns are also spreading to the infrastructure underpinning the AI industry. According to a recent Financial Times report, at least 47 major AI infrastructure projects have encountered setbacks during 2026 as financing difficulties, construction delays and electricity constraints disrupt expansion plans.
Microsoft also withdrew from a major AI data-center project in Essex, England, due to concerns about electricity supply.
Meanwhile, investors are scrutinizing the enormous amounts of capital required to build AI infrastructure, particularly as borrowing costs rise.
The developments do not necessarily signal the end of the AI boom. Instead, the trend suggests a shift toward financial discipline. Companies are now prioritizing efficiency, measurable productivity gains and returns on investment over aggressive spending.