AI sector needs to earn 6 trillion dollars per year to pay for data centers

The Global Technology Report from Bain & Company reveals that the AI sector must generate approximately 6 trillion dollars annually by 2031. This is necessary to cover the costs of global AI data centers, which are estimated to be around 1.5 trillion dollars per year. The required investment involves the construction of about 11,700 data centers. To cover these costs, revenue must be four times higher, according to the report.

Currently, AI applications, ranging from chatbot subscriptions to business software, generate about 1.8 trillion dollars in revenue. This results in a significant shortfall of 4.2 trillion dollars. Bain suggests that this added value must come from new markets that currently barely exist, which intensifies the challenge for the sector.

The annual bill for AI infrastructure will rise to 1.5 trillion dollars by 2031, mainly for new data centers and replacement costs for GPUs, memory, and networking equipment. The five largest cloud companies alone are set to invest 780 billion dollars in 2026, nearly five times the amount compared to previous years. Bain states that to achieve such investments, a revenue of 6 trillion dollars is necessary, with a quarter of this revenue needing to be reinvested.

David Crawford, chair of Bain's global technology practice, emphasizes the need for an annual global economic growth of about 1 percent point to make this financing sustainable. The remaining amount of 4.2 trillion dollars must come from yet untapped markets such as AI applications that replace search engines, autonomous transport, and physically advanced AI with robots. It is crucial that these new markets outgrow current developments in mobile technology and cloud solutions to alleviate the financial pressure on the sector.

Read the full article from AI Insider.