One of the genuine fears surrounding agentic AI is the question nobody likes to say out loud: What happens when it goes wrong? Or becomes unmanageable? We all know what AI can do. It can operate autonomously across enterprise systems, make decisions, and take actions without constant human approval.
But then Bill McDermott answered that directly on Mad Money Wednesday, July 22.
We have a kill switch that stops AI agents that go rogue.
Bill continued to tell Jim Cramer. “So those things don’t need to happen, and they wouldn’t happen, when companies run ServiceNow.”
That is a marketing statement. Yes. But it is also a product statement. And coming from the CEO of the company that has built the most widely deployed AI governance platform in enterprise software, it carries operational weight.
Also Read: ServiceNow Inc. Latest News and Stories
What the kill switch actually is and why it matters now
McDermott’s quote on Mad Money deserves to be read in full.
“The AI Control Tower is taking the Fortune 500 from AI chaos to AI discipline,” he said. “Token consumption has become a big issue. We manage that, enforce those policies, and the kinds of things that are happening in the media today around AI — we have a kill switch that stops AI agents that go rogue.”
The context is Uber’s situation, which I covered weeks ago. The company burned through its entire 2026 AI budget in four months and had to cap spending at $1,500 per employee per tool monthly.
More ServiceNow:
- BofA spots ServiceNow’s overlooked AI advantage
- Goldman says ServiceNow is writing a totally new playbook
- Bank of America resets ServiceNow stock price target sharply
Token consumption, agent drift, unauthorized access, and rogue workflow execution are real operational problems that enterprises encounter as they scale agentic AI deployments. ServiceNow’s AI Control Tower is positioned as the answer to that class of problem.
The Q2 business highlights show that the product is not theoretical. AI Control Tower expanded in Q2 with new discovery, observation, governance, security, and measurement capabilities across all AI systems.
Anthropic became the first design partner connecting Claude directly to ServiceNow via Action Fabric, according to the company’s earnings release.
ServiceNow’s Q2 2026 results and the AI revenue milestone that changes the narrative
The Q2 fiscal 2026 results, reported July 22, contained a number that I think the market has underreacted to.
ServiceNow AI crossed $1 billion in annual contract value in Q2 2026, according to the earnings release. That is a milestone that confirms the AI platform is generating real, contracted revenue at enterprise scale, not just deployment pilots.
- Total revenues were $3.987 billion, up 24% year-over-year (YOY)
- Subscription revenues of $3.877 billion grew 24.5% YOY.
- Current remaining performance obligations reached $13.20 billion, up 21% YOY
- Total remaining performance obligations stand at $29.0 billion.
“In an environment where most enterprises are still searching for AI’s ROI, ServiceNow is the platform delivering it,” said CFO Gina Mastantuono in the earnings release.
The company beat the high end of guidance across every topline and profitability metric for the second consecutive quarter, raising its full-year subscription revenue outlook.
The Q2 performance validates a business that is accelerating, not decelerating, despite a stock that is down 27.79% year-to-date and down 43.89% over the past year, according to Yahoo Finance.
ServiceNow AI crossed $1 billion in annual contract value in Q2 2026.
Bridget Bennett/Bloomberg via Getty Images
The long-term targets and partnerships actually frame the recovery case
The 22-year-old ServiceNow laid out its financial analyst day targets in Q2, too. By 2030, the company is targeting more than $30 billion in annual subscription revenues, with 30% of annual contract value driven by AI and a combined growth plus free cash flow margin exceeding 60%, according to ServiceNow disclosures.
The partnership ecosystem reinforces the governance positioning McDermott described. NVIDIA integrated AI Control Tower into its AI Factory design. Microsoft extended AI Control Tower governance across Microsoft Agent 365.
AWS surpassed $1 billion in Marketplace transactions with ServiceNow. Nearly all 50 U.S. states are using the ServiceNow AI Platform.
A look at the NOW technicals
Technically, NOW bottomed at approximately $81 in April 2026, well below the January 2025 all-time high near $239.
A monthly support zone at that April low, combined with an ascending trendline from that base, is what has driven the recovery to $110.
The next meaningful test is whether the Q2 earnings results and the AI Control Tower narrative can sustain that recovery with enough institutional conviction to push meaningfully higher.
McDermott’s kill switch comment is not just a soundbite. It is a product differentiation claim in the fastest-growing enterprise software category in history. I think that gives most people the confidence to keep on doing whatever advancement you’d want to achieve with AI.
For a stock that has given up more than 43% of its value over the past year while the underlying business accelerated to 24% revenue growth and $1 billion in AI ACV, that differentiation is the argument for why the gap between price and fundamentals eventually closes.