- The AI industry could need $6 trillion in yearly revenue, new report claims
- New AI products would need to generate trillions in commercial value
- Search, advertising, and autonomous systems could drive most future revenue
AI infrastructure spending is racing ahead, and by 2031 the industry will need to find far more commercial value to justify the bill, new research has claimed.
Bain & Company estimates yearly spending on facilities, processors, memory and networking could climb to $1.5 trillion by 2031 alone.
To keep that investment going, the industry would need roughly $6 trillion in annual revenue, assuming infrastructure swallows about 25% of sales every year.
Latest Videos FromTechRadar
Where would all that revenue come from?
Bain calls that 25% assumption bold yet fair, since it matches what cloud computing providers have spent on infrastructure in earlier years.
Even so, the math leaves AI businesses needing far more commercial activity than their current products bring in, so the hunt for new revenue begins.
The biggest piece, about $4.2 trillion, would come from new products in search, advertising, autonomous systems and physical AI, many of which barely exist today.
Enterprise productivity comes second, adding $1 trillion to $1.4 trillion as companies lean on AI for software development, sales, marketing, customer support and IT operations.
Consumer subscriptions and advertising trail far behind, contributing only $200 billion to $400 billion even as providers push AI products to billions of users.
Speed of adoption matters too, so leading AI labs are spending more than $9.75 billion on engineering that helps companies put AI to work faster.
“The economics of AI infrastructure demand trillions in new revenue beyond productivity gains,” said David Crawford, chairman of Bain’s global technology practice.
Crawford adds that the industry needs a surge of ideas big enough to dwarf what mobile technology and cloud computing unlocked in earlier years.
Beyond that, Bain says future products could reach into drug discovery, mental health and energy generation, areas where AI has limited presence today.
Data center expansion compounds the financial pressure
The physical infrastructure supporting AI is becoming larger and more expensive as companies continue increasing their computing capacity.
Bain says data center size and cost have been increasing at roughly double rates over periods lasting approximately 12 to 16 months globally.
According to Epoch AI, Meta’s Prometheus facility in Ohio had 600MW of capacity and an estimated $24 billion cost in 2025.
The facility could reach 2GW and $80 billion by 2027, before increasing to 5GW and $175 billion in 2029 under current projections.
By 2030, Epoch AI estimates the project could reach 9GW while requiring as much as $200 billion in total spending.
Those facilities require additional electricity generation, grid connections, advanced semiconductors, skilled workers and equipment capable of supporting sustained operations at scale.
The pace at which money is pouring into data centers has made securing enough computing capacity the industry’s most pressing concern.
“But the more important question may be whether enough economic value can be created to justify it,” the report said.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds.
https://cdn.mos.cms.futurecdn.net/ETt2Pe3AtJk2ynHY4vkd68-1920-80.png
Source link
Efosa Udinmwen
