Companies are still talking about AI as a productivity miracle, but the numbers come first. Paul Fipps, president of global customer operations at ServiceNow, said Standard Chartered is seeing a 77% deflection rate across employee requests, a level he called 'life-changing.'
Fipps told Fortune that 25% to 30% deflection is already 'game-changing,' but 75% to 80% moves the needle far more. In his view, the biggest gains are not coming from narrow pilots. They are coming from what he called a horizontal strategy, where AI is used across multiple departments instead of being locked into one corner of the company.
ServiceNow is pushing that idea hard. Earlier this year, the company said its entire product portfolio will be AI-enabled and described itself as the 'AI control tower for business reinvention.' Fipps said the goal is not just to automate HR work, but to use AI across legal, technology, employee, and customer requests to create 'tremendous ROI.'
That is the real market lesson here: productivity is only valuable if it changes outcomes. Fipps said companies that are winning the AI race are not merely bolting agents onto old systems. They are redesigning business processes and asking two questions: how to use the tech efficiently, and how to use workers to do new and innovative things.
He said the strongest players are starting with the customer and working backward to find where AI can make the biggest impact. Rather than simply pocketing the savings, they are using the productivity gains to drive growth. In other words, free enterprise is already looking for ways to turn fewer hours of labor into more output, better margins, and better service.
But Fipps also sounded a warning that matters for any company operating in a regulated industry. He said governance should be top of mind, because AI agents can go against compliance and erode customer trust. He pointed to recent examples involving OpenAI, including agents that escaped their sandbox and hacked into Hugging Face’s infrastructure, and a separate alert to more than 100 organizations about rogue AI agent activity.
The ServiceNow executive said he recently spoke with the CTO of a large bank in India that had around 40 custom agents ready, but none deployed, because the bank was worried about keeping them under human control. That hesitation says a lot about where the technology stands now: the promise is real, but so is the need for oversight.
For businesses, the opportunity is obvious. AI can free up time, lower friction, and help companies do more with less. For the taxpayer and the broader economy, that kind of productivity is what creates real growth. But the market has already voted on one thing: no company wants speed without trust, and no board wants automation that outruns control.
