ServiceNow, SAP and Workday are drawing new lines around the customer data stored inside their platforms as external artificial intelligence (AI) agents begin to erode the per-seat pricing model that has defined enterprise software for two decades.
Metered Layer Between Agents and Data
ServiceNow made the first move this week. At its Knowledge 2026 event in Las Vegas, the company unveiled Action Fabric, a new integration layer that external AI agents must pass through to access data and execute workflows inside its platform.
COO Amit Zavery said the company will meter that usage and charge customers for it, according to The Information. The pricing is action-based: customers pay according to how many operations an AI agent completes via the layer, a ServiceNow spokesperson confirmed. Anthropic’s Claude is the launch partner, with a connector that lets Claude Cowork plug directly into Action Fabric.
JPMorgan analyst Mark Murphy described the charge as effectively a tax on customers using outside AI agents to interact with data they already store in ServiceNow’s apps. Jon Sigler, ServiceNow’s executive vice president for its AI platform, pitched the product to Wall Street as a universal action layer that third-party systems would call into as a matter of course.
“What’s going to end up happening,” Sigler said, “is we’re going to have this universal action layer, where all of these systems are calling directly into our Action Fabric.”
The framing is consistent with where Workday has been heading. CEO Aneel Bhusri said earlier this year that charging for agent access offered considerable financial upside for the company.
Advertisement: Scroll to Continue
SAP Takes a Harder Lane
SAP is restricting access rather than metering it. The German enterprise software company updated its application programming interface (API) policy in April to prohibit third-party AI agents from interacting with its systems outside of SAP-endorsed architectures, according to The Register.
The company bars the use of SAP APIs for autonomous or generative AI systems that “plan, select, or execute sequences of API calls” without SAP’s official sanction. The company’s own AI assistant, Joule Agents, sits on the permitted side of that line.
According to CIO, the policy drew immediate pushback from the DSAG user group and from SAP’s own partner ecosystem, where vendors had built connectors into SAP data for tools including Microsoft Copilot and Salesforce Einstein. SAP CEO Christian Klein told investors customers would not pay to access their own data and that the company intends to keep its architecture open, per The Register. The policy language has not materially changed since he said so.
DataDog has gone a different route: capping how often customers can use its model context protocol server, which lets third-party AI agents interact with its products, to 5,000 daily requests or 50,000 monthly, with room for exceptions, as noted by The Information.
Revenue Under Pressure
The different approaches reflect a deeper tension. Traditional Software as a Service (SaaS) pricing worked because it mapped to people. Each license tied to an employee, a department and a cost center.
AI agents break that alignment. A single agent can trigger thousands of API calls in a day while adding no new seats. As PYMNTS reported, enterprise AI is replacing predictable per-seat billing with consumption models that behave less like subscriptions and more like utility invoices, leaving finance teams to manage spend that fluctuates with model activity rather than headcount.
That pressure is shaping how software companies respond. AWS CEO Matt Garman has publicly warned that incumbent software vendors that try to protect their installed base risk losing ground as AI reshapes how enterprises procure and deploy technology.
For ServiceNow and Workday, the tollgate strategy turns that disruption into a new revenue line. For SAP, a stricter posture on access attempts to route agent traffic through its own stack. Neither approach has been tested against enterprise buyers at scale.
PYMNTS tracked how Anthropic has begun charging enterprise customers based on their levels of AI use, a shift that is already straining AI budgets inside large companies. ServiceNow’s expanded AI Control Tower governance features are expected to reach general availability in August.