The artificial intelligence industry needs to generate $6 trillion in annual revenue by 2031 to justify the investments being made in data centers, according to a report from Bain and Company. The report, released on September 29, 2026, outlines that revenue from new product development is expected to be the largest contributor, projected to generate approximately $4.2 trillion over the next five years. This segment includes innovations in areas such as search, advertising, autonomy, and physical AI.
Enterprise productivity is estimated to require between $1 trillion and $1.4 trillion in revenue to support advancements in software development, sales, marketing, customer service, and IT operations. Bain's report highlights that the speed at which companies can implement AI, termed absorption speed, is becoming a critical competitive factor, with leading AI labs investing over $9.75 billion in engineering models to facilitate faster assimilation.
David Crawford, chairman of Bain’s global technology practice, stated, "The economics of AI infrastructure demand trillions in new revenue beyond productivity gains. What the industry needs is a wave of innovation that will dwarf what mobile and cloud unlocked."
Consumer-focused services, including subscriptions and advertising, are expected to contribute between $200 billion and $400 billion, which is considered essential as service providers aim to deliver AI-powered products to a global user base.
Bain forecasts that annual spending on AI infrastructure could reach $1.5 trillion by 2031, covering expenses for new facilities, increased capacity, and upgrades to existing GPU, memory, and networking equipment. If capital expenditures account for about 25% of industry revenue, sustaining this level of investment would necessitate an AI market nearing $6 trillion annually.
The report also notes that the rapid expansion of AI data centers is accelerating, with costs doubling approximately every 12 to 16 months. For example, Meta Platforms' Prometheus data center in Ohio is projected to increase its capacity from 600MW at an estimated cost of $24 billion in 2025 to as much as 9GW at a cost of $200 billion by 2030, according to Epoch AI.
Challenges for new data centers include increasing grid capacity, securing GPUs and other infrastructure components, and maintaining a skilled workforce. Additionally, public opinion and regulatory issues, such as concerns over resource use and noise pollution, must be addressed.
Governments in regions including the UAE, Saudi Arabia, the EU, South Korea, and the US are supporting the growth of the AI and data center industries, recognizing their importance for technological innovation, economic growth, and national sovereignty. Bain emphasizes that the capital needs for data infrastructure will remain high, with bottlenecks in power and semiconductors presenting further investment opportunities.