Agentic artificial intelligence could put as much as $234 billion in enterprise application software spending at risk by 2030, equivalent to about 20% of enterprise software-as-a-service expenditure, according to Gartner. The research firm said the shift would be driven by “agentic arbitrage,” in which AI agents perform tasks across several systems and reduce the need for employees to use individual software interfaces. This could weaken the relationship between user numbers and revenue for vendors that rely on seat-based licensing models.
“Agentic AI changes the economics of software,” said George Brocklehurst, managing vice president at Gartner. “Agentic systems deliver outcomes directly, bypassing traditional user experience-heavy applications and making the software invisible.” Gartner expects the development to change how enterprise software is designed, priced and used. The firm said SaaS products would not disappear but could be reorganised around automated workflows and business outcomes rather than dashboards and user access.
Enterprise buyers are also expected to place less emphasis on purchasing additional tools and features. Gartner said organisations would instead prioritise systems capable of retaining institutional knowledge and customer-specific context over time. Some software vendors have begun offering agentic products that can execute workflows, coordinate activity across systems and retain organisational knowledge. Gartner said these deployments currently tend to require extensive services support.
As adoption increases, software interfaces may become less important as a source of competitive differentiation. Existing vendors could lose market share to incumbent providers and new entrants offering horizontal platforms capable of operating across multiple enterprise applications. Gartner said established software companies would need to shift from interface-based products towards offerings tied more directly to outcomes. This could include embedding AI agents into operational processes and retaining customer-specific knowledge alongside underlying data.
The transition could create opportunities for AI-focused start-ups and service providers that provide an agentic layer across existing enterprise systems. Such companies could support cross-domain workflows, help organisations redesign processes and charge for measurable results rather than software features or user licences.
