Speed without intention gets you to the wrong destination faster. That’s what I’m seeing play out in conversations with customers across banking, healthcare, manufacturing—you name it. The organizations pulling ahead with AI share a common habit: Before racing to implement AI initiatives, they ask some deep questions.
The new Fortune AIQ 75 ranking, which draws on the ServiceNow Enterprise AI Maturity Index to evaluate how businesses are deploying AI relative to their peers, makes that visible.
In our Enterprise AI Maturity Index research, nearly every organization says it’s using AI, and 59% say they’ve moved beyond piloting. But only 9% have built genuinely autonomous, multistep workflows—automation that runs without a person rechecking every step.
In my experience working with organizations across industries, much of that 50-point disparity comes down to the questions AI leaders ask first. Here are the five questions I’d start with.
Before you ask how to make a process faster, determine whether it should happen in the first place. Automation can pay off most after a process has been pared down to the steps that still matter.
I recently worked with a financial services company that asked whether some of its processes were even necessary. Before deploying AI, the company standardized its knowledge base, documented every service process, and assigned an owner to every workflow.
Then it looked at the intake side for employee requests. Over time, more than 250 separate support queues had sprung up across departments, each one a reasonable answer to a specific query. Rather than trying to speed up 250 separate processes, however, the company asked: Which of these paths still need to exist?
As it turns out, most of them could be streamlined. The company dropped them into one front door for employee requests. Only then did the company layer AI on top. It’s now saving 2,200 hours per month, with 94% employee satisfaction—and it did so with a team of eight.
Many workflows are relics of an organization’s past, not blueprints for its future. AI deployed onto old assumptions just cements them.
Some of the most telling evidence that history isn’t destiny in this year’s Fortune AIQ 75 is in the industries people don’t think of when they picture an AI leader. The tech industry leads, with 17 spots on the list, and financial services is right behind it, with 12. But engineering and construction earned six spots. Healthcare also earned six. Food garnered five.
These are physical, asset-heavy, historically slower-moving industries that are proving they can turn AI into results, not just talk about it. Across 75 companies on the list, 18 sectors are represented. AI maturity isn’t concentrated in one corner of the economy anymore.
Many of the industries on this list carry decades of legacy systems and hard-won habits, yet they made the list. What sets them apart is a willingness to ask whether those norms still serve the business and to rebuild the workflows that don’t. It comes down to whether an organization is willing to build toward what it wants to be instead of defending what it used to be.
Technology can execute a flawed process flawlessly. That isn’t progress. The leaders getting this right treat AI as a design decision, not just a deployment.
A deployment question asks, “Where can we add AI?” A design question asks, “If I were building this today from scratch, with AI, what would I do differently?” Sometimes that means starting with a blank sheet of paper and sketching the workflow as if nothing exists. Then you add AI.
Adobe shows what the design mindset looks like in practice. When Toni Vanwinkle, Adobe’s vice president of digital employee experience, and her team applied agentic AI to the internal help desk that serves more than 30,000 employees, they didn’t stop at closing tickets faster. They also asked whether the answers were accurate, whether employees were satisfied, and whether the work gave capacity back to support staff.
Her test is the “so what?” that comes after the time saved. Using the ServiceNow AI Platform, Adobe’s IT and HR teams are also resolving cases 30% faster.
The AIQ 75’s top tier, the Pacesetters in the Enterprise AI Maturity Index, reports an average return on investment of 160% from their AI initiatives. And 36% report they’ve built autonomous, multistep workflows, compared with 2% of everyone else. The difference comes down to design and discipline.
Question 1 in this list is about whether a specific process should exist. This question is about whether the organization should keep backing existing priorities, including AI pilots and projects already underway.
Last year showed what happens when organizations fail to reassess their priorities. While everybody was talking about agentic AI, organizations’ AI maturity dropped by about 20% between 2024 and 2025, according to the Enterprise AI Maturity Index.
Here’s my hypothesis (which has held up in the customers I work with): too much noise, not enough leadership. Organizations took on more than they could carry, and attention and budget were spread thin. The businesses that matured were usually the ones that made tough decisions about what to prioritize and what to let go. In 2026, nearly every industry and geography rebounded.
Try the same test on your own portfolio. If an AI pilot or priority wouldn’t survive being proposed again today, the honest move is to end it, even if it has momentum and a sponsor. Stopping it can create room for the work that matters.
Disconnected systems and fragmented processes limit what AI can do, no matter how sophisticated the technology is.
Pacesetters have dealt with this: 64% say they’ve integrated their data digitally, compared with 14% of everyone else. And 57% say they share a vision for AI across the organization, compared with 21% of others. That’s the connective tissue.
This matters more now because the bar is moving. This year’s Enterprise AI Maturity Index added an assessment of organizations’ progress on the use of agentic AI, measuring the maturity, realized value, and scale of agentic deployments in production.
Just a few years ago, the question was whether an organization used AI at all. Now it’s whether AI agents are running and creating measurable value. In future years, the standard will likely shift again.
The businesses that lead are the ones that built the connective tissue beneath their AI initiatives early.
If you want to be on the Fortune AIQ 75 list next year, review this year’s takeaways. Then pick one workflow this week and answer all five questions honestly, including the uncomfortable ones.
Scale with intention. Intention must precede speed, every time.
Find out how ServiceNow can help you connect data, workflows, and AI on one governed platform.
