Built to bend: How AI-first supply chains adapt when disruption hits

An article from Sponsored Built to bend: How AI-first supply chains adapt when disruption hits Published June 22, 2026 Share Copy link Email / Print Permission granted by Zebra Technologies Sponsored content By Zebra Technologies SPONSORED CONTENT BY A product goes viral on a Tuesday.

By Friday, it’s sold out in three regions, overstocked in two others and gaining traction in places no one predicted. For planning teams tied to fixed schedules and calendar-driven cycles, that kind of speed has become the baseline. “The world is becoming more unpredictable,” said Nicholas Wegman, Ph. D.

, senior director and artificial intelligence scientist at Zebra Technologies. “And that unpredictability is being driven by lots of things — geopolitical forces, shifting consumer preferences, the way people shop and the expectations they have. All of that is changing so much that traditional ways of planning are struggling to keep up.”

The challenge, Wegman argued, extends beyond legacy technology. Many organizations are also running on outdated ways of working. When Cadence Becomes a Constraint Traditional planning workflows are built around a calendar: Teams gather on a set date, review inputs, approve a plan and move into the next cycle.

That rhythm can serve a stable business well, but as volatility increases, it becomes a liability. “If something is happening today, maybe I can’t wait until the next meeting or the next cycle,” Wegman said. “When you have a workflow based on a schedule or cadence that can’t quickly shift, that’s where things start to break down.”

That rigidity pushes teams into reactive mode — planners chasing explanations, updating spreadsheets or escalating decisions that could have been resolved sooner. In markets where a trend can appear and fade within a matter of weeks, detecting what matters and acting while the window is still open have become a core operating requirement.

Making Room for Higher-Value Decisions For Wegman, an AI-first workflow means artificial intelligence plays an active role in decisions — handling what falls within defined bounds and elevating the moments that genuinely need human judgment. “Think of AI as your junior planner or analyst,” he said. “There are certain things it can handle without you.

But it also needs to be smart enough to know what it can’t handle without you and say, ‘Here’s where I need a decision.’” Supply chain teams are being asked to do more with fewer people and less time. A well-designed system routes human effort toward the decisions that carry real weight while keeping routine work moving.

Reading the Signal Before the Moment Passes Demand intelligence — understanding what people are buying, what is driving that behavior and how long it is likely to last — is where AI can have some of its most immediate effect.

In industries where product life cycles are short and consumer behavior is volatile, the distinction between a sustainable trend and a one-time spike can determine whether an inventory strategy keeps pace. “The start of all of that is understanding what people really want,” he said.

“Using AI and machine learning to understand the drivers of demand and how that’s changing is the key to placing the right inventory and ordering what you need from suppliers.” Speed matters as much as insight. “Sometimes there’s a trend so short-lived that by the time it hits, you’ve missed it,” Wegman said.

“Other times, a trend hits and it’s a new growth category you need to build toward. AI and machine learning help you understand what the data is telling you — whether something is a one-off or needs to be incorporated into your everyday business.” The Human Layer AI-first supply chains still depend on experienced people.

The difference is that expertise gets applied where it has the most effect. Wegman described AI as providing a data-grounded starting point: something to respond to, refine and build on. In categories such as fashion and consumer goods, planners still bring deep market knowledge that no model can replicate. AI gives them a foundation.

“The predictions and recommendations that come out of AI are a first best attempt,” he said. “Then you can layer your judgment on top of that, grounded in sound statistical reality.” One of the most common missteps, Wegman noted, is dismissing recommendations simply because they differ from what a team would have done before.

“The point of AI is to tell you new things you don’t know,” he said. “It’s analyzing more data than any human can process, so it’s going to surface patterns that seem counterintuitive. If the AI is only telling you to do what you’ve always done, it’s not adding value.” He frames it as a bias check.

Humans naturally gravitate toward examples that reinforce existing assumptions. AI looks across a broader set of data. “The loudest voices in the room don’t always reflect what’s best for the business,” Wegman said. “AI can help organizations level-set because it doesn’t ha