ServiceNow’s AI Control Tower Flips the Script as Now Assist Surges Past $750M

When Nvidia’s Jensen Huang took the Computex stage and declared that AI agents represent an opportunity for enterprise software rather than a threat, the market’s bearish thesis on ServiceNow began to crack. The stock had already been hammered by fears of a “SaaSpocalypse” – the idea that autonomous agents would render per‑user licensing models obsolete and wipe out middleware platforms. Within 48 hours back in early February, the software sector had shed roughly $285 billion in value. ServiceNow, sitting in the uncomfortable middle layer between users and backend systems, looked particularly exposed.

That narrative has since undergone a full reversal. The key insight that emerged at the company’s Knowledge 2026 conference is that the more AI agents proliferate inside an enterprise, the greater the need for governance, orchestration and oversight. “Imagine there were no air traffic control and everyone just flew blindly,” one executive said, framing ServiceNow’s new “AI Control Tower” as the essential flight controller for the agentic era.

The AI Control Tower, unveiled at Knowledge 2026 alongside a unified interface called ServiceNow Otto, is a governance layer that discovers, monitors and steers every AI agent in an organisation – whether they are built on ServiceNow, powered by Claude, Copilot or home‑grown. The platform is model‑agnostic, the company stresses, and the real value sits in its proprietary workflow and data layers rather than any particular large language model.

That openness is already translating into hard numbers. Now Assist, ServiceNow’s generative AI product, reached an annual contract value of $750 million in the first quarter of 2026, up from $600 million at the end of 2025. The company promptly raised its full‑year target from $1 billion to at least $1.5 billion – a 50% upward revision that suggests it underestimated demand. Deals with Now Assist spending above $1 million grew 130% year‑on‑year, and 50% of new business is no longer priced on a per‑seat basis, validating the shift to a hybrid consumption model.

The financials paint a picture of accelerating momentum. First‑quarter revenue hit $3.77 billion, up 22% from a year earlier. ServiceNow now counts 630 customers with annual contract values above $5 million. For the second quarter, management guided for GAAP subscription revenue between $3.815 billion and $3.820 billion, while the full‑year forecast sits at $15.735 billion to $15.775 billion – growth of 22% to 22.5%. The long‑term target remains an annualised subscription revenue run rate above $30 billion by 2030, with AI contributing more than 30% of new contract value.

Should investors sell immediately? Or is it worth buying ServiceNow?

ServiceNow is also laying groundwork beyond its own platform. It became the first partner to integrate into Experian’s newly launched Agent Operating System, designed to help financial institutions scale AI agents securely. A multi‑year agreement with OpenAI gives customers direct access to frontier models including GPT‑5.2 without needing to build their own infrastructure. And in May, transactions through the Amazon Web Services Marketplace surpassed $1 billion.

The stock has reflected this repositioning with a volatile ride. After a 40% surge in just four sessions – fuelled largely by Huang’s remarks and the Knowledge 2026 presentations – shares have pulled back roughly 4% in the past week on profit‑taking. At around €103, the 30‑day gain still sits near 35%. The relative strength index reads 60.5, and the annualised 30‑day volatility is a hefty 74.5%, indicating that neither euphoria nor calm has taken hold. Insider activity bears watching: board member Teresa Briggs sold about $173,000 worth of shares on 1 June, and insider sales have outweighed purchases by $5.8 million over the past 12 months.

Analysts remain broadly bullish. The S&P Global consensus, drawn from 48 analysts, is a “Strong Buy” with an average price target of $141.86 – roughly 18% above current levels. William Blair’s Arjun Bhatia reaffirmed his buy rating on 1 June, arguing that the industry is pivoting from AI experimentation to production deployment.

ServiceNow’s structural edge may lie in its data moat. The platform processes 100 billion workflows and seven trillion transactions annually – a proprietary data set the company argues cannot be replicated through public open‑source sources. “That doesn’t exist in public open‑source data,” one executive noted.

The bear case is not extinguished. The risk of “shadow AI” – business units building agents without central IT oversight – remains real. And the possibility that future models bypass workflow orchestration entirely cannot be ruled out. But for now, the SaaSpocalypse shock has asked the right question at the wrong company. In a world overrun by agents, someone still needs to run the control tower. The market spent the first half of 2026 believing the opposite. It is now steadily revising that view.