Axios: AI infrastructure costs overtake workforce expenses

A senior Uber executive has stated that the company’s artificial intelligence budget for 2026 was exhausted significantly ahead of schedule due to the high cost of computing power and “tokens,” according to Axios.

The report notes that in some cases, spending on AI infrastructure is already surpassing payroll expenses, reflecting a broader rise in costs across the technology sector.

Global IT spending is projected to reach $6.31 trillion in 2026, an annual increase of 13.5%. The main drivers are said to be investments in artificial intelligence, cloud services, and software.

Industry participants confirm a shift in cost structure, where computing expenses increasingly outweigh personnel costs. A vice president at Nvidia responsible for applied deep learning, Bryan Catanzaro, said this trend has already become established. 

Similar dynamics are also emerging in startups, where companies increasingly speak about a model “scaling with intelligence, not headcount.”

Amid rising costs, investor pressure is intensifying, with demands for clearer proof of return on AI investments. Experts say companies will now need to demonstrate either efficiency gains or direct financial returns from these technologies.

By Jeyhun Aghazada