Agentic AI puts $234B in enterprise SaaS spending at risk, Gartner says

AI agents are poised to challenge traditional enterprise software business models, placing up to $234 billion in application software spending at risk by 2030 as they increasingly bypass human users and interact directly with business systems, according to Gartner.

“You are no longer buying software primarily for people; you are increasingly buying it for agents,” George Brocklehurst, managing vice president at Gartner, told CIO. “For a couple of decades, software has been evaluated on the interface, the user experience: usability, workflow, training. When AI agents become the primary user, all that depreciates.”

Gartner estimates that the exposed spending would account for about 20% of enterprise software-as-a-service (SaaS) spending by the end of the decade.

Gartner attributes the shift to what it calls “agentic arbitrage,” or the use of AI agents to complete business tasks across multiple enterprise systems, reducing the need for employees to interact directly with individual software interfaces.

Agentic AI changes the economics of software, Brocklehurst said, adding that these systems often bypass traditional software and deliver outcomes directly, breaking the link between user growth and revenue growth for many enterprise software vendors.

CIOs may need to rethink software procurement

The emergence of agentic AI will require CIOs to evaluate enterprise software differently, Brocklehurst said.

Instead of focusing primarily on user experience and interface design, organizations should assess whether AI agents can perform every business function through application programming interfaces (APIs) that human users can perform through application screens, he said.

“What really matters, as a starting point, is whether an agent can do everything—and more—through the system’s API that a human can do through a screen, and whether a vendor’s terms permit that,” he said.

That also changes how software contracts should be evaluated.

“Scrutinize the contract as much as you scrutinize the technology,” Brocklehurst said. “Vendors’ terms can prohibit or restrict — technically or financially — third-party autonomous use. CIOs may find their AI strategy blocked not by capability but by clauses they have already signed.”

He advised organizations to negotiate agent permissions into software agreements now because many existing contracts will remain in force when enterprise AI agents become mainstream.

Knowledge ownership becomes the next battleground

Beyond APIs and licensing, organizations should pay close attention to where AI systems retain operational learning, Brocklehurst said.

Every correction, exception, and workflow handled by an AI agent creates organizational knowledge, he said. Gartner refers to an organization’s ability to retain that knowledge as its Knowledge Retention Rate (KRR).

“If it accrues to the vendor’s shared models, your operational experience is improving a product your competitors also use,” Brocklehurst told CIO. “The most important clause in the next generation of software contracts is: ‘Who owns what the system learns from you?’”

According to Gartner, enterprises risk a new form of vendor lock-in if operational learning remains with software providers rather than the customer.

Traditional SaaS economics face disruption

According to Gartner, AI agents that execute work across multiple enterprise applications could reduce direct user interaction with traditional software interfaces, weakening the long-standing link between software usage and seat-based licensing.

Gartner said incumbent software providers should shift from interface-based value to outcome-based value, while embedding agentic capabilities directly into business processes and preserving customer-specific knowledge.

At the same time, AI-native startups and service providers could benefit by becoming the orchestration layer that coordinates work across multiple enterprise applications.

“While this shift is posing an existential threat for vendors who are defending legacy dashboards and seat-based models, it creates a substantial revenue opportunity for vendors who are enabling and developing services and platforms to support agentic-enabled cross-domain workflows,” Brocklehurst said.

Governance should evolve with autonomous systems

Gartner also urged CIOs to establish governance frameworks before autonomous AI agents become commonplace.

“Do not grant autonomy implicitly or unevenly,” Brocklehurst said. Organizations should treat agent autonomy as an explicit governance decision, defining where agents can operate independently, who authorizes those decisions, and how frequently those permissions should be reviewed.

“The companies that build that muscle now will move faster, and more safely, when the technology is ready for more,” he said.

Although Gartner described the transition as a redefinition of the long-discussed “Saaspocalypse,” Brocklehurst said SaaS itself would evolve rather than disappear. “This is less an apocalypse and more of a metamorphosis,” he said. “SaaS will not be destroyed; it will emerge in a different form.”