ServiceNow Reprices AI
ServiceNow is embedding AI across its platform and adjusting pricing to address enterprise ROI and budgeting concerns, signalling vendor moves to make AI features more cost‑manageable for customers. The change highlights how commercial packaging is forcing engineering teams to treat usage tiers and cost controls as architectural inputs. (techtarget.com)
ServiceNow is changing how it sells artificial intelligence after a year of asking companies to buy extra add-ons for tools like case summaries, chat replies, and search answers. On April 10, 2026, TechTarget reported that ServiceNow is now embedding more of those features across the platform instead of making every buyer start with a separate artificial intelligence purchase. (techtarget.com) ServiceNow’s software runs the back office for big companies: information technology tickets, human resources cases, customer support, asset records, and internal approvals. Its artificial intelligence layer, called Now Assist, writes summaries, drafts responses, answers search questions, and helps build workflows from text prompts. (servicenow.com) The old problem was not whether the tools worked. The problem was that finance teams had to guess the return before employees had used enough of them to prove the savings, and many Now Assist features were sold as paid upgrades on top of existing ServiceNow subscriptions. (techtarget.com) (servicenow.com) You can see that packaging in ServiceNow’s own pricing pages. Customer Service Management lists Now Assist for Customer Service “for an additional fee,” while higher tiers add things like Artificial Intelligence Agent Studio and Artificial Intelligence Agent Orchestrator on top of the base product. (servicenow.com) Its Information Technology Service Management page shows the same pattern in a different form. Foundation, Advanced, and Prime bundles each mix workflow features with different levels of Now Assist, platform artificial intelligence, and full artificial intelligence agents, which turns pricing into a design choice instead of a simple software purchase. (servicenow.com) That matters because ServiceNow has been moving from “copilot” tools that suggest text to “agent” tools that can take actions. In January 2025, the company introduced Artificial Intelligence Agent Orchestrator to coordinate teams of artificial intelligence agents, and in May 2025 it launched Artificial Intelligence Control Tower to govern models, prompts, datasets, and third-party agents from one place. (servicenow.com 1) (servicenow.com 2) Once software starts acting instead of just suggesting, companies want spending limits the way they want limits on cloud servers or text messages. ServiceNow’s own community examples show that one artificial intelligence task can consume 1 “assist” while another can consume 20, which means architecture teams have to think about quotas, routing, and guardrails before rollout. (servicenow.com) ServiceNow has room to make that trade because the company is already large and still growing fast. It reported 244 deals worth more than $1 million in new annual contract value in the fourth quarter of 2025 and said products including Now Assist outperformed during the quarter. (servicenow.com) So this is not a retreat from artificial intelligence. It is a sign that enterprise software vendors are learning that the hard part is no longer adding a model to the product, but fitting that model into annual budgets, approval cycles, and usage caps that a chief information officer can defend in a procurement meeting. (techtarget.com) (servicenow.com) The likely next step is that more software companies will hide basic artificial intelligence inside core tiers and reserve separate charges for heavy automation, autonomous agents, and governance tools. ServiceNow’s April 2026 pricing shift is an early example of that split: cheap enough to get adoption started, metered enough to keep the bill from surprising the buyer. (techtarget.com) (servicenow.com)