Winners and losers in AI adoption
Perspectives from BofA Global Research’s Leading Analysts
July 21, 2026
Tal Liani, Senior Research Analyst, Large-Cap, SMID-Cap and Security Software
AI: Creating winners and losers in enterprise software
AI is reshaping the software industry at an unprecedented pace. For some companies, it represents a fertile new growth frontier, unlocking demand, expanding addressable markets and creating entirely new monetization opportunities. For others, it introduces a disruptive force that threatens existing business models through disintermediation, commoditization and heightened competitive intensity. We view the Enterprise Software landscape through two distinct lenses: platforms and applications. While platforms, which include a range of offerings, from operating systems to IT management to e-commerce software and payments bundles, are positioned to emerge as AI beneficiaries, application providers face a greater risk of displacement. As with any technological disruption, AI is both constructive and destructive, creating new markets and revenue streams while simultaneously challenging incumbent solutions with more efficient, lower-cost alternatives.
Application vendors appear most exposed to AI-driven disruption.
By dramatically reducing the cost and complexity of software development, AI lowers barriers to entry, intensifies competitive pressures, accelerates commoditization and makes it increasingly difficult to sustain pricing power and incremental upsell opportunities for these narrower, more specialized offerings. In contrast, it’s our view that the long-term AI beneficiaries are companies that enable enterprises to deploy the technology safely, securely and at scale. These organizations serve as systems of record, controlling the data, workflows, governance and compliance frameworks that underpin successful AI adoption. We believe this dynamic favors vendors with broad product portfolios and platform solutions, as enterprises increasingly consolidate around trusted partners capable of delivering multiple AI-powered capabilities within a unified ecosystem rather than managing a fragmented collection of stand-alone applications.
AI Infrastructure: Demand continues to outpace supply
The proliferation of AI is also driving an unprecedented surge in infrastructure demand. Research laboratories, hyperscalers and emerging cloud providers require vast amounts of compute capacity, with demand for processing power continuing to outstrip available supply. Hyperscaler capital expenditures are expected to rise approximately 90% year over year in CY26 and grow to exceed $1 trillion annually in 2027, with spending heavily concentrated in AI-optimized data centers, power infrastructure, networking and Graphics Processing Unit deployment. While it is reasonable to expect the pace of investment to moderate over time, current spending trends remain exceptionally robust, with little evidence of a near-term slowdown. As hyperscalers and neocloud providers bring new capacity online, they start to convert their significant backlogs and remaining performance obligations (RPOs) into recognized revenue. These commitments have accumulated over the past several years as infrastructure build-outs constrained deployment timelines. The pace of backlog and RPO conversion could drive stock prices, as it would represent an important validation metric for the AI ecosystem, demonstrating both the commercial value of newly introduced AI services and the ability of cloud providers to translate extraordinary capital investment into sustained revenue growth.
AI buildout pumps up oil & gas needs
Pace of activity suggests that behind-the-meter power is evolving from a niche solution into a major pipeline to meet data center energy demand.
