The White House names more than 40 trading partners as transshipment risks and deploys a pre-clearance artificial intelligence screening system, shifting customs enforcement from post-release audit to the moment of entry
WASHINGTON, August 16, 2026. The White House Office of Trade and Manufacturing Policy has released a 25-page report accusing Chinese exporters of routing goods through more than 40 countries to disguise their origin and evade United States tariffs, and paired it with an artificial intelligence screening system designed to score shipments for transshipment risk before Customs and Border Protection makes a release decision.
The report, titled “The Great Transshipment Scam” and published Thursday, August 13, sorts the flagged economies into three tiers and estimates annual revenue loss to the Treasury at between $19 billion and $26 billion. It names longstanding allies alongside transit hubs, and it makes transshipment compliance an explicit condition of ongoing reciprocal trade negotiations.
For importers, the consequential change is not the country list. It is the timing of enforcement. The system, called Detective Border, is designed to produce a risk score before goods are released into commerce, rather than during a post-release audit that historically could take years to resolve. That inverts the practical economics of origin misdeclaration: an importer who previously could clear customs, sell the goods, book the margin and litigate the assessment later now faces the possibility of a hold, an examination and a denial of entry before there is anything to sell.
The three tiers and what they signify
The report groups flagged economies by scale, degree of economic integration with China, and what it terms “weak-link advantages.”
Tier 1, labeled “Diversified Scale Leaders,” names Canada, the European Union, India, Israel, Japan, Mexico, South Korea and Taiwan. The White House was careful to characterize this designation as a function of trade volume rather than government complicity. The argument is that very large legitimate flows provide statistical cover within which illicit flows are difficult to isolate. That framing did not fully defuse the diplomatic difficulty of listing treaty allies in the same document as higher-risk transit jurisdictions.
Tier 2, “Scale Leaders with Significant Economic Integration,” includes Brazil, Indonesia, Malaysia, Thailand, Turkey and Vietnam. These are the economies that absorbed the largest share of the manufacturing relocation that followed the 2018 tariff round, and where Chinese-owned production capacity is most heavily represented.
Tier 3 covers a broader roster across Latin America, the Middle East and elsewhere, including Panama, Colombia, Costa Rica, Argentina, Chile, Peru and the Dominican Republic. These are identified principally as transit hubs with elevated routing risk rather than manufacturing centers.
Senior Counselor Peter Navarro told reporters Thursday that countries enabling Chinese goods to circumvent United States tariffs will face consequences, and that the transshipment issue “is going to be part of the reciprocal trade negotiations.” Treasury Secretary Scott Bessent framed the initiative in enforcement terms, saying the administration is pairing trade policy with enforcement to ensure goods entering the country pay the duties they owe.
China’s Embassy in Washington responded that any unilateral actions or agreements concerning transshipped goods must not target or harm the interests of third parties.
What Detective Border does
The system is better described as a fusion of several analytical capabilities than as a single product.
According to the report’s enforcement architecture section, Detective Border integrates shipment data against routing histories, verifies production capacity and ownership relationships, and analyzes packaging patterns and port X-ray imaging to detect mismatches between what a container declares and what it contains.
The technical foundation predates the announcement. In October 2025, CBP awarded a multimillion-dollar contract to Exiger, a McLean, Virginia supply chain analytics firm, specifically to modernize transshipment detection. Exiger also holds a position on a $919 million blanket purchase agreement with the General Services Administration covering supply chain risk illumination tools.
Three analytical layers do the work. The first is entity resolution: the platform reconciles records across many data sources into an ownership graph spanning hundreds of millions of organizations, which allows a Vietnamese-registered shipper to be traced through intermediate holding structures to an ultimate Chinese beneficial owner. The second is anomaly detection: the system cross-references billions of shipment records against production capacity data by country and product category, flagging cases where a transit country’s declared exports of a good exceed its plausible capacity to manufacture that good. The third is imaging analysis: CBP’s existing non-intrusive inspection infrastructure generates tens of thousands of scan images daily, and models trained to detect discrepancies between declared manifests and imaged cargo now process them in near real time.
The output is a composite score reflecting whether a container’s routing, ownership structure, production plausibility and physical imaging all cohere with its declared origin, delivered before the release decision rather than after.
CBP Commissioner Rodney Scott, who testified at his April 2025 confirmation hearing that the agency needed major investment in artificial intelligence for non-intrusive inspection, has described the initiative as a fundamental shift in enforcement approach. He also testified that artificial intelligence would augment rather than replace officer judgment, and the system as described preserves human decision authority over final clearance. Administration officials are targeting full-scale implementation by the end of 2026.
The numbers, and the disagreement about them
The report assembles five estimates of the scale of illegal transshipment, from Goldman Sachs, the White House Council of Economic Advisers, Exiger, the Commerce Department’s Office of Trade and Economic Analysis, and the data analytics firm Altana. They span a range from $40 billion to $303 billion annually, and the report acknowledges they are not directly comparable because they use different methodologies and definitions.
Exiger’s figure, a central estimate of $75 billion in illegally transshipped goods for the period from February 2025 to February 2026, anchors the administration’s public messaging and corresponds to the $19 billion to $26 billion annual revenue loss headline.
The Commerce Department contributed what appears to be the most methodologically conservative figure. Its test looked only at cases where the identical product, at the eight-digit Harmonized Tariff Schedule level, was imported from China and then re-exported to the United States from the same region in the same quarter. Restricted to Mexico, India and Vietnam, that narrow test identified approximately $67 billion in transshipped goods last year, with roughly $28 billion in associated lost revenue. The administration’s summary is that “the direction is the same,” even where the magnitudes differ.
Critics have concentrated on precisely that methodological spread. Jiang Wenran, founding director of the China Institute at the University of Alberta, told China Daily that the report blames other countries for a problem created by tariffs the Supreme Court subsequently ruled illegal, and that the $75 billion centerpiece is a model-derived estimate rather than a verified loss.
William Reinsch, senior adviser at the Center for Strategic and International Studies and a longtime analyst of the issue, has framed the underlying difficulty in terms enforcement officials would likely accept: “Nobody admits they’re illegally transshipping,” which puts the burden on customs authorities in both transit countries and the United States to identify it and assess the additional duties. Reinsch describes that as a hard task, particularly in countries with close commercial ties to China and little incentive to assist American customs officers.
The report also supplies one figure that is not an estimate at all. CBP data cited in the document show that comparing 526-day windows before and after January 2025, shipments flagged with post-release discrepancies rose 245 percent, and associated revenue assessments climbed from $9.6 billion to $25.8 billion. Whatever the true size of the underlying evasion, the enforcement environment has already changed materially.
The line between diversification and evasion
The hardest problem in this initiative is not catching deliberate fraud. It is that legitimate supply chain diversification and illegal transshipment can look similar in the data.
The “China Plus One” strategy that manufacturers pursued in good faith after 2018, shifting portions of final assembly to Vietnam, Thailand, Mexico and elsewhere, frequently produces structures that resemble origin laundering on paper. The legal test is substantial transformation. Under 19 C.F.R. section 134.1(b), processing must produce a genuinely new article with a new name, character or use. Assembly alone does not satisfy it.
That standard has real teeth in electronics. A laptop whose motherboard, display panel, battery pack and chassis all arrive from China and are assembled in Vietnam generally remains Chinese-origin for customs purposes. Vietnam’s electronics sector imports components worth nearly as much as it exports in finished goods, with Chinese-origin inputs accounting for something on the order of 5 to 20 percent of total product value depending on segment. Firms whose Vietnamese or Mexican operations consist principally of final assembly of Chinese components therefore carry genuine legal exposure even where no one intended to deceive anyone.
Fortune, reporting on the Detective Border rollout, noted that illegal transshipment is notoriously hard to detect, that origin determination is complex when a product incorporates components from multiple countries, and that it is unclear how much of the supply chain shift the report describes reflects illicit activity rather than legitimate changes in global production.
Detective Border’s production-capacity cross-referencing is designed to isolate the outliers, cases where declared exports structurally exceed documented manufacturing capacity. But the report itself acknowledges that effective enforcement requires distinguishing genuine manufacturing and substantial transformation from pass-through trade and origin shifting, and that distinction is a judgment call that a risk score can inform but cannot make.
Compounding exposure
Two other elements of the enforcement architecture magnify the consequences of a flag.
The first is Executive Order 14411, signed June 3, 2026, which overhauled the rules governing importers of record. The order mandates minimum bond coverage tied to import volume, requires beneficial ownership disclosure for every importer, bars foreign importers of record from filing informal entries, establishes a “good standing” gate capable of excluding a non-compliant importer from the United States market entirely, and sets a penalty floor of at least 50 percent of any assessed penalty with no mitigation for repeat offenders. CBP is implementing the requirements across 45, 90 and 180 day windows, with full compliance required by November 30, 2026.
The second is retroactivity. Navarro said Thursday that importers found to have falsified a product’s country of origin can face tariffs applied retroactively for approximately one year of prior shipments. Combined with the penalty floor, the exposure for a company with a year of poorly documented entries through a flagged jurisdiction is not one container’s duty. It is every container moved over the preceding twelve months at the rate that should have applied, plus at least half again in penalties.
The report also notes that the recently concluded Bilateral Agreements on Reciprocal Trade contain provisions allowing signatories to set rules of origin specifically designed to prevent preferential treatment from extending to third-country goods. The practical effect is that a country which negotiated a favorable rate with Washington now carries an affirmative obligation to police its own customs administration, or risk losing the preference it bargained for. Navarro’s characterization of the Tier 1 list as a warning rather than an accusation follows that logic: cooperate on enforcement before transshipment becomes a formal negotiating issue.
Sector exposure
The burden of the new screening posture does not fall evenly. Four sectors carry concentrated risk.
Electronics and consumer technology sit at the center. The category combines high tariff differentials between Chinese and non-Chinese origin, complex multi-country bills of materials, and assembly-heavy operations in exactly the Tier 2 jurisdictions the report highlights. It is also the category where the substantial transformation analysis is most contestable, because the highest-value components are frequently the ones that never leave the Chinese supply chain.
Solar and energy storage face a similar structure with a longer enforcement history. Origin scrutiny in that sector predates this initiative by several years, and the administration added a further layer on August 6 with a framework establishing minimum import prices for several polysilicon goods. Companies in that supply chain should assume they are already inside the screening perimeter.
Furniture, textiles and apparel carry a different profile: lower unit values, very high transaction counts, and margins thin enough that examination delay and detention costs matter as much as duty exposure. For these importers the operational risk of a flag can exceed the financial risk of the underlying assessment.
Automotive and industrial parts distribution sits at the intersection of both problems, with large SKU counts, complex sourcing and duty rates high enough that origin determination changes the economics of the product.
Sectors with simpler supply chains, single-country sourcing and clear manufacturing footprints face materially less exposure, and companies in that position should be able to document it quickly if asked.
How this fits the wider August picture
The transshipment initiative is one component of an unusually concentrated stretch of United States trade activity, and reading it in isolation understates its significance.
Within the same week, the Court of International Trade upheld the elimination of the de minimis exemption, closing the low-value parcel channel permanently and pushing that volume into the ordinary entry system where Detective Border operates. President Trump signed a Section 232 proclamation imposing tariffs of up to 100 percent on imported drones and components. Commerce’s proposal to sweep 14 additional categories of steel, aluminum and copper derivative products into Section 232 remains open for comment until August 27. And negotiations with Canada continued against an August 19 deadline for 50 percent duties on roughly $20 billion in Canadian goods, with autos, dairy and alcohol at the center of the talks.
Those actions rest on different statutory authorities and are administered by different agencies, but they compound in the same place: the importer’s entry file. A single shipment can now carry a Section 232 sectoral duty, a Section 301 exposure, an origin challenge under the transshipment initiative and an entry obligation created by the end of de minimis. Each is administered on its own clock, and none of the agencies involved coordinates its deadlines with the others.
That compounding is itself an argument for centralizing trade compliance rather than distributing it across procurement, logistics and finance, which is how it sits in many mid-sized importers. The companies that have handled 2026 well have generally been the ones with a single owner of tariff exposure and a current, auditable product master.
What importers should do now
The advisory consensus from trade counsel since June has been consistent, and Thursday’s report should be read as the final notice for companies that have not acted on it.
Audit origin documentation before CBP initiates the review, not after. That means bills of materials tracing every input to its country of origin, factory production records evidencing the specific manufacturing operations performed in the transit country, value-added calculations showing what share of total product value was created locally, and certificates of origin capable of surviving a field audit. Companies that did this work during the Vietnam factory enforcement actions in late July are better positioned than those that did not.
Map exposure by jurisdiction. Any sourcing route passing through a Tier 1, Tier 2 or Tier 3 country now carries elevated screening probability. That is not a reason to abandon those routes, most of which are entirely legitimate, but it is a reason to know which SKUs travel which paths and to have the substantiation ready for each.
Test the substantial transformation position on the products where it is closest. Where an operation consists principally of assembly, the honest internal answer may be that the goods are Chinese-origin, and the cheaper course is to correct the declaration prospectively and consider a prior disclosure rather than wait for a flag. Prior disclosure remains available and materially limits penalty exposure, and Executive Order 14411’s penalty floor makes that calculus more favorable than it was a year ago.
Build delay into logistics planning. Even a flagged shipment that ultimately clears absorbs examination time, demurrage and detention. Companies operating on thin inventory buffers through a peak season already strained by record ocean rates should assume longer and more variable clearance times on affected lanes.
Confirm importer of record arrangements against Executive Order 14411 before the November 30 deadline. Arrangements in which a foreign supplier acts as importer of record are directly affected, and the bond and beneficial ownership requirements take time to satisfy.
Finally, revisit contractual duty allocation. Where a supplier warrants origin, that warranty should be backed by an indemnity that covers retroactive assessment and penalties, not merely the duty differential on the shipment in question.
Assessment
The policy case for the initiative is not hard to state. If tariffs are the instrument, evasion undermines them, and origin fraud disadvantages exactly the domestic producers the tariffs are meant to protect. Enforcement at the moment of entry is more effective than enforcement years later, and the technology to do it now exists.
The difficulties are equally clear. The revenue estimates span nearly an order of magnitude, which invites the criticism that the number was chosen to fit the policy rather than the reverse. The country tiering names close allies in a document about evasion, which carries diplomatic cost that the careful framing only partly offsets. And the fundamental analytical problem, distinguishing substantial transformation from pass-through, is a legal judgment that no risk score resolves.
The practical exposure for American businesses is this: a supply chain that passed scrutiny under 2024 enforcement standards may not pass under 2026 standards with identical manufacturing and identical documentation, not because anything changed in the factory but because the system evaluating it now sees the ownership graph, the routing history and the capacity mismatch simultaneously, and sees them before the goods are released.