The rapid build-out of artificial intelligence infrastructure will require about US$6 trillion in annual revenue by 2031 to justify the capital being deployed, with much of that value needing to come from new products and services beyond conventional employee productivity gains, according to Bain & Company.
In its seventh annual Global Technology Report, Bain said existing AI applications in consumer subscriptions, advertising and enterprise use cases such as software development, sales, marketing, customer service and IT operations could generate between US$1.2 trillion and US$1.8 trillion in annual revenue.
That still leaves about US$4.2 trillion that would need to come from new categories of innovation.
Bain identified four areas with the potential to fill that gap: AI model providers monetising search-like experiences and advertising; autonomous systems spanning vehicles, trucks and drones; physical AI such as digital twins, simulations and robotics; and entirely new applications in areas including drug discovery, mental health and energy.
“The debate today is fixated on employee productivity. The economics of AI infrastructure demand trillions in new revenue beyond productivity gains,” said David Crawford, chairman of Bain’s global Technology practice.
He added that AI infrastructure is being built ahead of the demand curve and that funding it sustainably would require adding roughly 1% to annual global GDP growth.
Bain said the AI boom has also reversed the long-running dominance of software over hardware.
From 2020 to 2026, hardware and semiconductor stocks grew at a 24% compound annual rate, compared with 6% for software.
Among the fastest-growing segments are high-bandwidth memory, advanced packaging and custom silicon, including application-specific integrated circuits.
The report said hyperscalers and AI-native companies are increasingly designing chips around specific workloads such as training, inference and agentic computing, pushing custom accelerators further into the mainstream.
At the same time, capacity constraints could emerge in more conventional memory products as manufacturers prioritise HBM investment, potentially contributing to tighter supply and higher prices for smartphones and PCs.
Supply-chain strategy is also becoming a competitive differentiator as companies respond to geopolitical risks, export controls and natural-disaster exposure by diversifying suppliers and production locations.
Bain said AI is reshaping cybersecurity as well.
The time required to execute a typical cyberattack has compressed from an estimated four weeks to around 18 hours, while the proliferation of AI agents is broadening potential attack surfaces.
The report said leading companies are increasing remediation budgets by double-digit percentages and reallocating as much as 20% to 25% of cybersecurity personnel towards handling alerts generated by AI-powered scans.
Bain also highlighted growing concern over how third-party vendors deploy AI throughout the life of a contract, particularly as software and models are updated continuously.
Another emerging differentiator is what Bain calls “AI absorption speed” — how quickly companies can translate AI capability into business use.
Leading AI labs are investing as much as US$9.75 billion in forward-deployed engineering models designed to help enterprise customers adopt and integrate AI more quickly.
Bain does not expect large language models to become fully commoditised. Instead, it sees a market segmented between frontier models for complex and high-value tasks and lower-cost, specialised models for more mature and routine use cases.
The report also found that senior technology leaders expect dramatic improvements from AI in software development over the next one to two years, including a 148% improvement in release-cycle speed and a 95% uplift in developer productivity.
However, Bain said current gains remain much lower, at roughly 20% to 27% across key measures.
It warned that faster coding alone does not guarantee faster software delivery because bottlenecks can shift into review, coordination, quality assurance and governance.
Bain said companies that capture the most value will be those that redesign the full software development lifecycle, structure knowledge effectively for AI agents and embed quality and trust into workflows.
