Author: Naomi Chan

Trade disruption, pandemic-era supply failures, and geopolitical fragmentation have made local-for-local manufacturing a strategic priority. Understanding what sovereignty investment actually costs — and delivers — is now essential.

The Strait of Hormuz carries 20% of global oil and all of Qatar’s LNG. Iran’s April 2026 tanker seizure pushed Brent above $105. The risk is permanent — and most businesses aren’t managing it.

The Taiwan Strait carries 90% of the world’s advanced semiconductor production past a credible military threat. Every manufacturer has exposure to what happens there — most don’t yet manage it explicitly.

India is being positioned as the world’s next manufacturing hub. The headline narrative is bullish. The granular reality — on infrastructure, skills, and regulation — is more complicated. A hard look.

The US has deployed Section 301 tariffs against 16 countries at once — the broadest trade action in history. For multinationals managing supply chains and pricing, the rules have fundamentally changed.

As enterprises race to deploy AI agents across their operations, they are inadvertently creating a new class of security vulnerability that conventional cybersecurity frameworks were not designed to address. The hidden cost of AI deployment is not just compute and talent — it is an expanded attack surface that grows with every agent you add.

NVIDIA controls roughly 80 per cent of the AI accelerator market — a concentration that is now prompting the world’s largest technology companies to build their own silicon. The shift from monolithic GPU clusters to modular chiplet architectures is not a distant forecast: it is already reshaping enterprise procurement, investment strategy, and the competitive dynamics of AI at scale.

AI sovereignty has moved from the margins of technology policy into the boardroom agenda of every serious multinational. With 93 per cent of global executives calling it mission-critical and at least 34 countries enacting data localisation rules, the question is no longer whether to build a sovereign AI strategy — but how fast.

Global semiconductor revenue is set to exceed $1.3 trillion in 2026 — the highest growth rate in two decades — as DRAM prices surge 125% and NAND flash climbs 234% year-on-year. For business leaders outside the technology sector, the chip super-cycle is no longer a supply chain footnote: it is a balance-sheet emergency hiding in plain sight.

Global capital is quietly migrating away from US-dominated markets. Dollar uncertainty, geopolitical fragmentation, and emerging market maturation are driving the largest reallocation in a generation.