The global AI boom will need to generate about $6 trillion in annual revenue by 2031 to justify the massive investment in infrastructure needed to power the technology, according to Bain & Company’s seventh annual Global Technology Report.
While existing consumer and enterprise AI applications could generate up to $1.8 trillion in annual revenue, Bain says the industry will need to create another $4.2 trillion through entirely new products, services and markets.
AI’s insatiable compute demand would require $6 trillion in annual revenue by 2031 and much of the value lies in new innovation, beyond employee productivity, the report said.
Existing applications of AI will grow. Consumer AI, through subscriptions and advertisements, and enterprise AI, through software development, sales, marketing, customer service, and IT operations, could total between $1.2 trillion and $1.8 trillion in revenue.
In addition to these applications, new categories of innovation are required. Bain’s research finds four key categories that are likely to fund the remaining $4.2 trillion of new revenue.
* First, model providers are replacing search engines and integrating ads to generate new revenue.
* Second, autonomous everything particularly in automobiles, trucks, and drones, as well as other industrial automation, will create new products and services.
* Third, physical AI, including simulations, digital twins, and robotics, will unlock a wide range of new applications in R&D and manufacturing.
* And last, new products and uses that don’t exist today will enable new markets and opportunities from abundant intelligence – these may include drug discovery, mental health, and energy generation.
“The debate today is fixated on employee productivity. 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. AI infrastructure is being built well ahead of the demand curve and funding it sustainably will require adding approximately 1% to the annual global GDP growth rate,” said David Crawford, chairman of Bain’s global Technology practice.
Hardware strikes back
The massive demand for AI compute has revived the hardware industry. In a reversal, hardware and semiconductor stocks grew at a 24% compound annual rate from 2020 to 2026 vs. 6% for software. Three of the fastest growing segments include high-bandwidth memory (HBM), advanced packaging and custom silicon, with application-specific integrated circuits (ASICs) scaling rapidly.
The co-development of dynamic random-access memory (DRAM) with logic silicon means switching is more difficult. And with major players focusing on HBM capacity, there is little investment in double data rate and NAND, which could worsen shortages and raise the prices of smartphones and PCs.
Special-purpose, high-performance accelerators are moving from niche to mainstream as hyperscalers and AI-native firms design chips tuned to their workloads. As a result, custom chips are grabbing a larger share of the data center compute market. The turning point is the emergence of grand-scale, homogeneous workloads (training, inference, and agentic) that now run well past the volume needed to amortize a custom design.
Pricing and supply risks from natural disasters, geopolitical disruption, and export controls are rewriting supply chain dynamics. Companies are now securing multiple supply locations with diversified vendors, while foundries are responding similarly, particularly in logic chips.
"This is a dynamic time for players in the hardware sector. Supply chain and procurement strategies are increasingly sources of competitive advantage as investing in supplier capacity, long-term agreements, equity investments in the supply chain, and multi-vendor, multi-geography sourcing are now very high on the C-suite agenda. Also, product strategy choices are changing. The old model where one vendor innovates and sells to everyone else is changing, and a wave of verticalization and semi-custom design are becoming more prevalent,” said Anne Hoecker, global head of Bain's Technology practice.
Cybersecurity faces watershed moment
In recent times, high-profile incidents, notably involving frontier AI model tests, have put cybersecurity at the front and centre of every CISO’s agenda. AI has massively compressed the time required by a typical cyberattack from an estimated four weeks to about 18 hours, Bain finds, and the proliferation of AI agents expands its impact. Poor agentic housekeeping is a glaring weakness with an understandable cause: The lack of a silver-bullet solution from vendors is leading many firms to hesitate when they should be proactively creating a pragmatic and flexible solution through a build/buy/partner approach.
The Bain CISO survey also highlighted supply chain risk as an issue. Companies must understand and control how vendors deploy AI through the whole life of a contract, including midcycle changes and fourth-party exposure. But with vendors shipping changes to models and software daily, oversight systems based on infrequent questionnaires aren’t coping.
Leading companies are strengthening their remediation operating model. Leaders have increased remediation budgets, typically by a double-digit percentage, while redirecting as much as 20% to 25% of their cybersecurity human resources from other work to remediate alerts from AI-powered scans, the report finds. DevOps teams are also pushed to find and mitigate vulnerabilities. Leaders are also prioritising action in the most concentrated and hard-to-fix AI risks, including those posed by legacy platforms, network layers and software-as-a-service vendors.
AI absorption speed becomes new competitive advantage
Absorption speed, the pace at which companies can put AI to work, has become the new competitive variable, Bain reports. To address this, leading labs are investing upwards of $9.75 billion in forward-deployed engineering models to help companies assimilate faster. Vendors are also building out the application and infrastructure layers which act as harnesses that connect AI to the enterprise and turn model intelligence into business outcomes.
Even as model competition rises and token prices fall, Bain does not see a commoditisation of large language models.
“We see the more likely scenario as a continuum of frontier and mature models. New and unproven cases will initially favour frontier models, then likely migrate to lower-cost alternatives as the cases mature. Frontier providers will push toward more fundamental challenges, capturing value where superior intelligence matters most. Meanwhile, lower cost, specialized, and optimized models will efficiently serve more common tasks, providing the services and support that enterprise customers need,” said Crawford.
“The industry end state is far from settled. We are likely headed into a segmentation between frontier models and trailing models or an expanding definition of models rather than a classic commoditisation pattern.”
As companies become more adept at deploying the right model for the right task, another challenge surfaces – building the engineering management systems that make those tools effective. Bain’s latest Tech and Engineering Survey of 293 senior technology leaders found that respondents anticipated a 148% improvement in release-cycle speed and a 95% uplift in software developer productivity within the next one to two years. That’s much higher than the productivity gains most are capturing today: between 20% and 27% across key metrics.
While AI speeds coding, it shifts bottlenecks into review, coordination, quality and governance, so optimising individual tasks without redesigning the full software development lifecycle moves friction rather than improving end-to-end throughput. Leading companies adopt three strategies: they have structured the knowledge on which AI agents depend, they engineer quality and trust into the workflow, and they manage the software development lifecycle (SDLC) as a continuously improving product. -TradeArabia News Service
