By Our Correspondent
National News – Artificial Intelligence (AI) adoption is driving operating costs beyond employee salaries in several global technology firms, including Nvidia, Uber, and emerging startups, according to industry executives and reports on Friday, May 8, 2026.
The development highlights how compute-intensive AI tools, coding assistants, and automated agents are reshaping corporate budgets worldwide.
Nvidia Vice-President of Applied Deep Learning, Bryan Catanzaro, explained that the cost of compute for his team now exceeds staff salaries due to heavy usage of AI infrastructure.
Uber’s Chief Technology Officer, Praveen Naga, also admitted that the company underestimated its AI budget as token-based pricing rapidly increased expenditure.
Industry observers say the shift from fixed software licensing to token-based pricing is accelerating unpredictability in enterprise AI spending.
Nvidia CEO Jensen Huang reportedly encouraged greater AI tool usage, suggesting engineers earning high salaries could justify substantial annual AI token spending.
The issue, which affects firms in the United States and beyond, shows that AI services are billed per usage rather than fixed subscription models.
This means costs rise sharply when companies run continuous AI-driven coding, data processing, and workflow automation.
A startup example revealed that Swan AI, a four-person company, received a $113,000 monthly bill from Anthropic, illustrating how small teams can face enterprise-level expenses, highlighting uneven economics of AI adoption for small teams.
Analysts also note that AI agents already contribute to about 11% of live code updates at Uber, signaling deeper automation trends.
A 2024 MIT study cited in reports found humans performed better in 77% of tasks compared with AI systems in similar conditions.
Despite rising costs, companies continue to invest heavily in AI, viewing it as a long-term productivity tool.
However, experts warn that firms without clear budgeting strategies risk financial strain as automation expands faster than cost efficiency improvements.
