- ServiceNow, SAP, and Workday to charge for external AI agent access.
- Shifts from per-seat licenses to metered, per-API-call pricing models.
- Workday faces pressure as AI threatens its per-employee billing model.
Back

Enterprise software giants are erecting a new paywall for artificial intelligence, shifting from per-seat licenses to a metered access model that could add billions in new revenue. The move, led by ServiceNow, SAP, and Workday, effectively creates a toll booth for external AI agents that need to access the vast stores of customer data held within their platforms.
"The first wave of AR software relied on rigid, rule-based systems that still required so much manual intervention," said Rocio Wu, Partner at F-Prime, in a statement regarding a related investment. "With the rapid advancements in LLM and agent capabilities, Fazeshift is going after a massive, largely untapped opportunity with true automation."
The strategic shift comes as AI agents begin to handle complex enterprise tasks, threatening the traditional per-user pricing that has defined the software industry for decades. For Workday, the change is particularly acute. The company's stock is already trading at a 43 percent discount to the sector's price-to-earnings average as investors question the durability of its per-employee billing model in an era of AI automation, according to a recent Seeking Alpha analysis. The report noted Workday's FY27 guidance signals decelerating growth and rising capital expenditures.
This new pricing strategy positions the enterprise software platforms as the central governance layer for AI-driven work, a market also being pursued by data platform companies like Snowflake, Databricks, and Teradata. By charging per API call, these firms are betting they can capture value from AI-driven activity regardless of where the AI agent itself originates, turning a potential threat into a significant revenue stream.
The move toward metered access reflects a broader industry transition from AI copilots to autonomous agents that execute end-to-end workflows. Startups like Fazeshift, which recently raised $22 million, are building AI agents to automate entire accounts receivable processes, operating across ERPs, CRMs, and email systems. These agents rely on the data locked inside platforms from SAP, Workday, and others to function. By implementing a pay-to-play model, the incumbent software providers are ensuring they get a cut of this new, automated activity.
Teradata's recent launch of its Autonomous Knowledge Platform highlights the competitive push to become the operating layer for enterprise AI. The platform is designed to help companies govern AI agents' access to data and the actions they can take. This focus on a governed "control plane" is central to the strategy of the major software players, who argue that their platforms are the natural place to enforce security, compliance, and auditability for AI-driven actions. The question for customers will be whether this new toll is a fair price for governance or a tax on innovation.
This article is for informational purposes only and does not constitute investment advice.