At a glance

• The global economy is increasingly defined by two forces: geopolitics and AI.
• AI is not just software—it increasingly resembles infrastructure.
• The value AI creates may be substantial, but who captures that value remains uncertain.
• The response may begin—and end—with reindustrialization.

Geopolitical frictions – most recently the conflict in the Gulf, but before that the rewiring of the global trade and tariff landscape – are challenging the model of supply chains built for high efficiency, low latency and low redundancy.

AI, meanwhile, promises a productivity upside, but the timeline and magnitude of that promise remain uncertain. The build-out is well under way: chips, data centres, power grids, cooling systems, talent and capital are being mobilised on an unprecedented scale.

These two forces each have distinct implications for the global economy. On growth, they pull in opposite directions. Yet both require capital, and both stress supply chains and energy systems.

How governments respond to this pivotal moment will determine the future of business and society the world over. Just as we are reminded every day of industrial revolutions of the past, resilient reindustrialisation must define our future.

Weaponising Interdependence

To look at equities markets, one would be tempted to conclude that geopolitics scarcely mattered. But this picture can be misleading: the companies at the centre of the AI boom are still booming, they are driving indices, and they can withstand the effects of even the US-Iran conflict. Many others cannot.

While the worst of the disruption to energy supplies appears to be behind us, the blockade of the Strait of Hormuz has shown what the weaponisation of interdependence means for supply chains, growth and inflation.

Nowhere is this more apparent than in Asia, where energy systems and manufacturing sectors continue to rely on oil and gas imports. Twenty-four governments in Asia and the Pacific have introduced energy conservation measures since the start of the Iran conflict.¹ Consumer prices have raced higher, most notably in the Philippines and Vietnam.²

But while geopolitics is testing the resilience of the old global system, new technologies are testing whether we can build a new one.

Repeating, or Rhyming?

And so we come to AI, and the current moment. The first industrial revolution brought class struggle, Marxism, and then labour laws. The second brought the Great Depression and then the social security system. The computer revolution brought the dot-com boom and bust, but much less institutional reset.

The lesson is that institutions lag. Technology moves first, then capital. Society catches up last of all.

Will AI be different? Perhaps its capital intensity sets it apart. It’s clearly not just software. In many ways it looks like infrastructure: it needs chips, data centres, power, cooling, grids, land, talent and capital. The current capex boom is exciting because it shows conviction, but dangerous because investment can run ahead of adoption and valuations ahead of cash flows.

Where AI is similar to the leaps of the past is that adoption will take time. Economy-wide transformation requires companies and governments to change processes, incentives, governance, compliance, skills and organisational habits. Those changes take place at the speed of humans, not silicon.

What we don’t yet know is how the gains will be shared out. Will they be evenly spread between the hyperscaler that supplies the compute, the manufacturer that raises its margin, the worker who feels supercharged, and the customer who pays less?

It is far from clear today that whatever gains do transpire will be distributed evenly or predictably at all.

The Way to Respond

Our response to the twin challenges of geopolitics and AI is complicated by the classic macro tensions of growth and inflation.

In the current system, both of these forces have the potential to cause another inflationary shock and derail global growth. While energy prices have retreated from their April peaks, the Hormuz saga has shown how quickly geopolitical tensions can spill over to the real economy. Capex spending on AI continues to increase. Central banks clearly understand this, which is why they have shifted to a more hawkish stance in recent meetings.

We need a new system – one that is capable of withstanding geopolitical frictions and capturing the benefits of AI. The response, then, begins and ends with reindustrialisation.

The resurgence of geopolitical tensions has focused policymakers’ attention on reducing interdependence and building economic resilience. In the AI era, national security means not just armies, but infrastructure, manufacturing, agriculture, data centres, energy systems and critical minerals. All of these today can be disrupted by a small number of global chokepoints.

What this means for Western economies is hard to overstate. Their reindustrialisation effort must be massive, and it must involve rewriting industrial policies and rethinking incentives.

For decades, the Western model has relied on consumers, services and asset markets. But the new model needs factories, grids, industrial finance and state coordination. Investments will necessarily tilt towards industries that are secure, scalable and strategically aligned.

As economies seek to balance today’s technological hopes and geopolitical anxieties, that new model must no longer deliver only efficiency, but also resilience.

References

<sub>1. https://www.iea.org/data-and-statistics/data-tools/2026-energy-crisis-policy-response-tracker 
2. https://asia.nikkei.com/economy/inflation/asean-inflation-vietnam-and-philippines-hit-hardest-by-iran-war</sub>