AI and Supply Chain Resilience: A New Era of Economic Power

For years, supply chain management was considered important but unremarkable. The pandemic, trade wars, export controls, sanctions, climate shocks, and competition over AI, energy, and critical minerals have changed this. At Davos, it was clear that supply chains are now central to economic power and a significant source of systemic risk.

This was the central theme in my discussion with Brandon Daniels, CEO of Exiger. The company has spent the past decade using advanced AI to answer a key question: where does risk reside in the global supply chain?

“Supply chains win wars,” Daniels said early in our conversation. Without resilient, transparent supply chains, he argued, you don’t get data centers, you don’t get energy transitions, and you don’t get AI at scale.

Risk everywhere, visibility nowhere

What’s changed isn’t just the volume of disruption—it’s the density. Daniels described a world in which risk now comes from every direction at once: natural disasters, artificial crises, tariff regimes, sanctions, and regulatory volatility. He cited estimates that global supply chain disruptions have already cost the global economy more than a trillion dollars this year alone.

That aligns with the broader data. The World Trade Organization reports that the number of trade-restrictive measures imposed by governments has increased more than fivefold since 2015, with export controls and industrial policy now a permanent feature of global commerce. The World Bank and OECD have repeatedly warned that fragmentation—not globalization—is now the dominant trend shaping trade and production.

Against that backdrop, Daniels argues that resilience isn’t about reacting faster—it’s about planning further ahead. “Not five weeks out, not five months out,” he said, “but five years out.”

That shift—from just-in-time efficiency to long-horizon resilience—was one of the quiet but consistent themes across Davos this year.

The myth of the visible supply chain

One of Daniels’ more unsettling observations is how little visibility most companies actually have into what they depend on. “Most people think a company has all of the information to manage their supply chain at their fingertips,” he said. “The fact is, they don’t.”

In industries like automotive manufacturing, Daniels noted that the overwhelming majority of production costs come from goods and services sourced outside the company’s own walls—often from suppliers several tiers removed, operating in facilities the buyer doesn’t own, audit, or even know exist.

Supply chains don’t look like pyramids, he said. They look like diamonds—wide at the top, briefly diversified in the middle, and then narrowing down to a handful of irreplaceable choke points. “One company in the world” that makes a specific piece of photolithography equipment. A dozen firms are capable of producing high-purity quartz crucibles for silicon wafers. Lose one node, and entire industries stall.

AI, in this context, isn’t about prediction for its own sake. It’s about exposing those hidden dependencies before they become failures.

Compliance is now existential

Regulation was once something supply chains adapted to. Now it actively shapes them.Daniels pointed to the proliferation of multi-tier compliance obligations since the pandemic—U.S. export controls, European due diligence regimes, and laws such as the Uyghur Forced Labor Prevention Act. “They’ve skyrocketed,” he said, and companies no longer have the luxury of treating compliance as a box-checking exercise.

That’s especially true because the ethical and operational risks tend to cluster. Forced labor, Daniels argued, “runs in packs” with environmental abuse and unsafe industrial practices.

The numbers are sobering. According to the International Labour Organization, roughly 50 million people worldwide are trapped in modern slavery today. Daniels described how forced labor is embedded upstream in industries most consumers never see: chemicals, steel alloys, critical minerals, pharmaceuticals, and semiconductor inputs.

In pharmaceuticals alone, he warned that a meaningful share of active pharmaceutical ingredients are produced in lightly regulated or unregulated facilities, creating risks that are ethical, environmental, and ultimately medical.

In response, Exiger recently released a portion of its AI-generated forced-labor mapping data publicly, working with organizations including Hope for Justice, the Slave-Free Alliance, and leading academic researchers. The goal, Daniels said, is to force visibility where plausible deniability has long thrived.

Reshoring, but selectively

If globalization is fracturing, what replaces it? Not isolation—but prioritization.

Daniels doesn’t see a wholesale retreat from global trade. Instead, he expects governments to identify a narrow set of truly critical materials and capabilities—semiconductors, energy systems, shipbuilding, advanced manufacturing—and ensure domestic or allied access to those assets.

Automation changes the calculus. Manufacturing may not return jobs at scale, but it can restore capacity, boost GDP, and reduce exposure to geopolitical shocks. What emerges, Daniels suggested, will be “alliances of specialization,” where countries double down on what they do best while drawing hard lines around national security.

Transparency becomes the price of openness. “You want to make sure that chip doesn’t end up in a missile,” he said. That requires independent monitoring, shared standards, and systems that can track risk across borders—not merely trust it.

The Davos takeaway

If Davos 2026 had a quiet consensus, it was this: an AI strategy without a supply-chain strategy is illusory. Models may scale quickly, but minerals, energy, labor, and logistics do not. The constraints are physical, ethical, and political—and they’re tightening.

The companies and countries that win the next decade won’t just build better algorithms. They’ll build supply chains that can withstand a world defined by friction.