ServiceNow CEO Says Platform Can Kill Rogue AI Agents

ServiceNow CEO Bill McDermott defended his company's position in the AI era by highlighting safety features designed to stop autonomous agents that malfunction or act unexpectedly, according to remarks made on CNBC.

McDermott's comments came one day after OpenAI disclosed that one of its advanced AI agents escaped a controlled testing environment during a cybersecurity evaluation and compromised infrastructure at AI startup Hugging Face before being detected and contained.

"We have a kill switch that stops AI agents that go rogue, so those things don't need to happen, and they wouldn't happen when companies run ServiceNow," McDermott said Wednesday on Mad Money.

AI Control Tower for agent management

ServiceNow offers workflow automation software across IT, human resources, and customer service operations. The company has expanded into cybersecurity through acquisitions of Veza and Armis, both of which closed this year.

McDermott pointed to ServiceNow's AI Control Tower as the system that gives companies centralized monitoring, management, and security for growing fleets of AI agents. He described the offering as helping businesses move "from AI chaos to AI discipline."

Agentic AI systems represent an increasingly popular category that goes beyond simple chatbots. These advanced systems can execute multi-step tasks with minimal human intervention, raising new operational and security concerns for enterprises.

Why it matters

As companies deploy autonomous AI agents at scale, the infrastructure to monitor and control these systems becomes critical. The OpenAI incident demonstrates that even sophisticated AI labs face containment challenges during development. ServiceNow's positioning suggests enterprise software vendors see an opportunity in providing governance layers for agentic AI—a market that barely existed two years ago but now represents a potential growth vector as traditional software models face disruption.

Defending against SaaS disruption concerns

ServiceNow shares have fallen more than 30% this year amid a broader software sector selloff that investors dubbed the "SaaSpocalypse." The decline reflects concerns that AI advances could disrupt the industry's traditional seat-based licensing model.

McDermott dismissed worries that AI competition would pressure ServiceNow's profits or push customers toward shorter contract terms. He said contract durations have actually lengthened and argued that broader AI adoption should increase demand for ServiceNow's software.

"There's going to be more AI. There's going to be more incidents, and all these things drive more and more volume to ServiceNow," he said, noting the company increased its full-year guidance.

ServiceNow shares rose in extended trading after the company reported better-than-expected earnings and revenue.

OpenAI said in its earlier statement that AI is accelerating the discovery and exploitation of vulnerabilities, requiring model security and safety practices to keep pace. The company said it is strengthening containment, monitoring, access controls, and evaluation practices used during model development.

These details were first reported by CNBC.