Enterprise AI's $4.7tn growth prize remains largely unclaimed

Bain analysis puts 75% of AI's decade-long profit opportunity in innovation and market-share shifts, not the efficiency gains most firms are chasing.

By Disrupts Editorial Team Sun, 18 October 2026 AI level 2 of 5

AI level 2 of 5: drafted by our AI editorial assistant from the company's announcement; fact-checked, edited and signed off by Mark Walker, Editorial Director. What the levels mean

A brightly lit data center aisle extends between two rows of server racks displaying glowing green and blue status lights visible through their panels.

Enterprise technology leaders have spent the past three years measuring artificial intelligence against a single metric: cost reduction. New analysis from Bain & Company suggests that framing may be costing them the bigger prize.

Bain puts $4.7 trillion of global business profit at stake from AI between 2025 and 2035. Of that total, only $1.1 trillion derives from productivity gains. The remaining share breaks down into $2.2 trillion from innovation and $1.3 trillion from competitive market-share shifts. In Bain's own framing, roughly 75% of the opportunity lies beyond efficiency. Meanwhile, Gartner's September 2026 forecast places worldwide AI spending at $2.67 trillion this year, up 49.5% year on year, with spending on AI agents and assistants projected to more than double again by 2027 to $65.5 billion.

The efficiency trap

The gap between where the value sits and where organisations are looking is striking. McKinsey's State of AI survey, which polled 1,719 participants across 97 nations between May and June 2026, found approximately 80% of organisations pursuing efficiency gains from AI. Only 6% of those McKinsey classifies as high performers were also deploying AI in pursuit of growth or innovation. UK data from the Office for National Statistics reinforces the pattern: improving existing business operations is the most common AI application, cited by over 60% of larger British businesses, while developing new offerings and exploring new markets remains a minority pursuit.

UK AI company SEPANTA, which positions itself as an operating system for enterprise growth, says it is seeing the shift begin in its order book. The company reports signing seven new enterprise contracts in a single month, including one six-figure agreement, with clients drawn from global consumer goods businesses. It argues that what corporate teams are now requesting is less about headcount reduction and more about unlocking bandwidth for growth. CEO Omid Pakseresht puts it plainly: "Efficiency is the floor, growth is the point."

The convergence read-across

The strategic implications extend well beyond enterprise software. If the centre of gravity in AI investment does shift from cost-cutting to revenue generation, the downstream effects will ripple across sectors. Capital markets have largely priced AI infrastructure spending on the premise of productivity-driven ROI, which is a metric visible on cost lines. A transition to growth-oriented deployments, where returns are diffuse, longer-cycle and harder to audit, will test how institutional investors model AI exposure across their portfolios.

For cross-sector leaders, the more immediate question is organisational architecture. SEPANTA's CPTO and co-founder Dr Nima Pakseresht identifies the core problem as fragmentation: most large organisations have accumulated a portfolio of AI tools that do not communicate with each other, creating gaps where value leaks out between systems. This is not a software problem unique to any one vertical. A global logistics operator, a pharmaceutical company running parallel clinical and commercial operations, or a financial institution managing siloed data estates all face the same structural issue: AI tools that optimise individual functions while leaving cross-functional opportunity invisible.

SEPANTA has financed its growth without institutional venture capital, generating over £1m in revenue through customer contracts, angel investors, and three crowdfunding rounds. That bootstrapped trajectory is itself a signal worth noting: at a moment when VC appetite for early-stage AI has become more selective, the company's ability to grow on customer revenue alone suggests enterprise buyers are willing to pay for AI that demonstrably connects fragmented systems, not merely automate within them.

The broader market is still in early innings on this transition. Whether the Bain projection proves accurate will depend substantially on how quickly enterprise buyers can shift procurement criteria from efficiency metrics they already know how to measure to growth metrics they are still learning to define.