AI was supposed to be the threat hanging over enterprise software. This week, ServiceNow (NYSE:NOW) became one of the clearest examples of a different argument.
The stock rallied as investors rotated back into software names after weeks of concern that AI agents could reduce the need for traditional SaaS platforms. Nvidia CEO Jensen Huang helped shift the mood, arguing that AI agents are likely to use more software tools, not fewer, and pointing to established software companies, including ServiceNow, as potential beneficiaries of the agentic AI cycle.
The move does not settle the debate. It does show that investors are starting to separate software that may be commoditized by AI from software that may become more important because AI needs a trusted system of record, workflow and permissions layer to operate inside large companies.
The workflow layer is back in focus
ServiceNow is not just another application sitting on an employee’s desktop. Its platform is used across IT service management, HR, customer service, security operations and internal workflows. That gives it a role inside the operating system of a large enterprise.
That matters in an AI world.
AI agents can summarize, write, classify and recommend. But inside a real company, they also need to know who is allowed to approve a request, which record should be updated, what policy applies and where a task should move next. Those are workflow questions, not chatbot questions.
In plain English, AI still needs somewhere to do the work safely.
That is the part of the ServiceNow story investors appear to be rethinking. If AI agents become more common across enterprises, demand may shift toward platforms that can connect those agents to business processes without breaking controls, compliance or audit trails.
Now Assist gives the AI story a product line
ServiceNow has been pushing that argument through Now Assist, its generative AI layer across the Now Platform. The company says Now Assist is designed to improve productivity through conversational and proactive experiences, with AI capabilities tied into workflows, records and enterprise processes.
The early financial signs have helped the case. In its first-quarter 2026 results, ServiceNow reported US$3.67 billion in subscription revenue, up 22% year over year. Current remaining performance obligations reached US$12.64 billion, up 22.5%, while total remaining performance obligations rose 25% to US$27.7 billion. The company also said customers spending more than US$1 million in annual contract value on Now Assist grew more than 130% year over year.
That is the important distinction. The AI story is not just a slide in the investor deck. It is beginning to show up in customer spending, at least among larger accounts.
ServiceNow has also laid out a longer-term target of more than US$30 billion in annual subscription revenue by 2030, roughly double its projected 2026 level, with AI expected to become a larger part of annual contract value over time.
The valuation still needs proof
The rally has a clear logic. ServiceNow has recurring revenue, deep enterprise relationships and a product suite that can expand across departments. If AI increases the amount of work running through enterprise systems, the company may benefit from higher usage, premium AI features and broader platform adoption.
That is the constructive case now being priced back into the stock.
The risk is that the market may move faster than the evidence. ServiceNow remains a premium software name, and premium multiples usually require clean execution. Investors have already shown sensitivity to margin pressure, including after the company’s first-quarter update, when concerns around acquisition-related costs weighed on the stock despite strong revenue growth.
The next test is simple: AI enthusiasm has to keep turning into contract expansion.
Investors will likely be watching Now Assist adoption, large-customer spending, current remaining performance obligations and management commentary on AI pricing. They will also be watching whether AI improves ServiceNow’s own margins over time, or whether product investment and acquisitions keep pressure on profitability.
For now, ServiceNow has put a sharper question in front of the market.
AI may still disrupt parts of software. But for platforms that control workflow, identity, approvals and enterprise records, the story may be more complicated. ServiceNow does not need investors to believe AI is harmless. It needs them to believe AI agents still need a place to work.
That is the debate now sitting inside the share price.
