The surge in generative AI has prompted widespread debate about whether traditional software business models are at risk. A new long-range assessment from Bernstein provides a five-plus year perspective that frames the technology as a market expander rather than an existential threat to the sector.
Cloud infrastructure set to gain
Bernstein's report singles out the foundational layers of the technology stack - primarily Infrastructure-as-a-Service (IaaS) and Platform-as-a-Service (PaaS) - as the areas most advantaged by the generative AI wave. The firm anticipates that major hyperscalers will capture substantial growth as enterprises increase demand for both GPU/ASIC and CPU capacity to handle the increasingly complex requirements of model training and inference.
The research highlights that this trend will be reinforced as "agentic AI" - autonomous, goal-oriented software agents - becomes more common in enterprise settings. Such agentic capabilities, Bernstein says, will require robust, scalable back-end support, further strengthening demand for cloud-based infrastructure.
PaaS and databases: evolution in usage
Within the PaaS layer, Bernstein expects a notable shift in how enterprises leverage platform functionalities. As database usage grows under AI-driven workloads, the firm projects that legacy on-premise database systems will keep migrating to the cloud. This movement is forecast to create room for specialized, AI-native database solutions that better serve compute-heavy, model-centric applications.
Bernstein's five-plus year outlook notes that while nimble, new vendors may win leadership positions in certain narrowly defined niches, these gains are not interpreted as a wholesale replacement of incumbent software vendors. Instead, the integration of generative AI is expected to produce additional layers of value across the software stack.
Investor takeaways
Investors, according to the report, should prioritize companies that can bridge traditional software requirements with the new, compute-intensive needs of an AI-driven, agentic infrastructure. The research argues that the competitive landscape will change, but that the underlying demand for scalable, cloud-based architecture is likely to remain strong for the remainder of the decade.
Overall, Bernstein concludes that the narrative suggesting the "death of software" is overstated. Rather than signalling the end of established software businesses, generative AI appears poised to broaden the total addressable market for cloud infrastructure and AI-native platform services.
Limitations and scope
The report focuses on forward-looking changes through 2030 and emphasizes structural demand drivers for cloud capacity, PaaS evolution, and database migration. It notes potential shifts in vendor leadership in specific niches, while maintaining that the broader software ecosystem is fundamentally resilient.