The numbers come first: AI supply chain firm Exiger estimated $75 billion in illegally transshipped goods moved through third countries between February 2025 and February 2026, representing a tariff revenue loss the White House pegs at between $19 billion and $34 billion.
The Trump administration is now deploying what it calls an AI-powered 'detective border' to close that gap. The system will scan shipment data against routing histories, confirm production capacity and ownership relationships, and analyze packaging patterns and X-ray imaging at ports to detect mismatches between declared cargo and actual contents.
The legal and analytical framework comes from a report released Thursday by the White House Office of Trade and Manufacturing Policy, led by senior counselor for trade and manufacturing Peter Navarro. The report accuses more than 40 countries of participating in China's 'shadow transshipment network,' ranking them 'according to the scale of China-linked trade, the depth of their economic integration with China, and the weak-link advantages that make them susceptible to rerouting activity.'
'This is basically a warning to the world — don't try to cheat America,' Peter Navarro told Bloomberg Television on Thursday.
The report's named enablers span the globe. On U.S. land borders: Mexico and Canada. Among major economies: the European Union, India, Japan, and South Korea. Other flagged nations include Indonesia, Thailand, Brazil, and Malaysia — each cited for comparative advantages that can be exploited, including low labor costs, strategic port access, lax customs enforcement, or free-trade zones.
The report is candid about the incentive structure. 'Local firms capture assembly fees, warehousing revenue, logistics margins, port charges, customs brokerage income, land rents, and export-processing-zone investment,' it states. 'Governments benefit from jobs, tax receipts, foreign investment, and trade growth.' In other words, transshipment pays — until enforcement catches up.
The trend accelerated after Trump's first-term tariffs on Chinese exports prompted the so-called China +1 strategy, pushing manufacturing into Vietnam, Cambodia, and other countries, often through Chinese-owned factories. The report acknowledges that 'tariff differentials' created by country-specific duties increase the incentive to illegally transship, but argues that the new AI tools are designed to 'detect, deter and prevent' evasion at scale.
U.S. trade enforcement has historically concentrated limited resources on the highest-payout cases, which are complex and can take years to build — allowing illicit goods to continue undercutting American producers in the meantime. The report itself notes the core challenge: 'Effective enforcement therefore requires distinguishing legitimate manufacturing and substantial transformation from pass-through trade and origin shifting.'
The figures driving this crackdown draw on two government and three private-sector sources, with Exiger's $75 billion figure representing the mid-range estimate.
CEO Times take: Tariff policy only works if it is enforced. A $19-to-$34 billion annual revenue gap is not a rounding error — it is a structural subsidy to foreign intermediaries and Chinese exporters at the expense of compliant American importers and the U.S. taxpayer. Deploying AI at the border is the right instinct: the administrative state has long lacked the bandwidth to police origin fraud at volume. If the 'detective border' delivers on its design, it will be one of the few cases where technology and trade policy reinforce each other in a way that actually protects free enterprise rather than distorting it.
