// Field Intelligence · Forced Labor
The US Franchised Its Forced-Labor Import Ban. Your UFLPA Program Won’t Travel.
The United States priced a foreign statute, collected, and is now distributing the implementation manual. For importers, compliance just stopped being a single-jurisdiction question — and the jurisdictions don’t agree on who bears the burden of proof.
Six countries rewrote their import laws in seven weeks.
Between USTR’s publication of proposed Section 301 action on June 5, 2026 and final action on July 23, Cambodia, Guatemala, Honduras, India, Sri Lanka, and Trinidad and Tobago each adopted a forced labor import prohibition. Jordan undertook a commitment to do so in its Agreement on Reciprocal Trade. In each case, that move is the stated basis for a 10 percent tariff rather than 12.5.
On September 15, USTR convened representatives of more than 50 trading partners and taught them how to build and enforce such a prohibition, working with the Department of Homeland Security, U.S. Customs and Border Protection, and the Department of Labor. Country-specific technical assistance accompanies the training.
Together those facts describe something new in U.S. trade policy. The United States priced a foreign statute, collected, and is now distributing the implementation manual. For importers, the consequence is not a new U.S. obligation. It is that the compliance question has stopped being a single-jurisdiction question, and the jurisdictions do not agree on who bears the burden of proof.
// The mechanics, briefly
At the President’s direction, USTR initiated 60 Section 301 investigations on March 12, 2026 into trading partners’ failure to impose and effectively enforce prohibitions on importing goods produced with forced labor. The covered economies account for 99.4 percent of U.S. imports. On June 2, USTR determined the conduct was unreasonable, burdened or restricted U.S. commerce, and was actionable under Section 301(b), finding that 54 economies had failed to impose and effectively enforce a prohibition and six had failed to effectively enforce one.
Final action issued July 23, published in the Federal Register on July 28, and applied to goods entered for consumption, or withdrawn from warehouse, on or after 12:01 a.m. eastern time on July 24, with a narrow in-transit exception running to July 28.
The rate structure is the instrument:
- 10 percent for economies that impose a prohibition, have committed to impose and enforce one through an Agreement on Reciprocal Trade, or maintain a partial regime that prevents importation of certain forced labor goods — Argentina, Bangladesh, Cambodia, Canada, Ecuador, El Salvador, Guatemala, Honduras, India, Indonesia, Jordan, Malaysia, Mexico, Pakistan, Sri Lanka, Trinidad and Tobago, and the United Kingdom.
- 10 or 12.5 percent net of MFN for certain non-exempt products of the European Union, Taiwan, Japan, Korea, and Switzerland.
- 12.5 percent for all others.
Two carve-outs matter more than their placement in the notice suggests. All articles and parts of articles subject to Section 232 tariffs are exempt, so a derivative already carrying 232 duty does not stack a forced labor 301 on top. Separately, USTR was directed to establish tariff-rate quotas for Bangladesh, Cambodia, Indonesia, and Malaysia, with an initial three-year duration, sized to each economy’s purchases of U.S. cotton and textile goods, letting a set volume of textiles and apparel enter free of the 301. Those TRQs do not yet exist. Until USTR finds establishment feasible and publishes an effective date, 10 percent applies to the covered lines. Anyone modeling Southeast Asian apparel landed cost through 2027 is modeling an unpublished quota.
// Four regimes, four burdens of proof
This is where a single global traceability file stops working.
United States. The Uyghur Forced Labor Prevention Act applies a rebuttable presumption. Goods mined, produced, or manufactured wholly or in part in Xinjiang, or by a listed entity, are presumed made with forced labor and prohibited under 19 U.S.C. § 1307. The importer carries the burden, must answer CBP’s information requests, and must demonstrate by clear and convincing evidence that forced labor was not involved. The Entity List reached 187 after DHS added 43 companies on July 31, effective August 3, and CBP’s dashboard reflects more than 18,000 shipments worth over $3.8 billion reviewed since June 2022. Behind UFLPA sits Section 307 of the Tariff Act of 1930, whose consumptive demand exception Congress repealed in 2016.
European Union. Regulation (EU) 2024/3015 applies from December 14, 2027, and rests on the opposite premise. There is no blanket presumption. Competent authorities must establish forced labor through a risk-based investigation beginning from a substantiated concern, and the burden rests with the authority rather than the economic operator, whose duty is to cooperate and produce evidence on request. Proposals to import a U.S.-style presumption for high-risk regions were considered and rejected during negotiation.
Canada. Tariff item 9897.00.00 has prohibited goods produced wholly or in part by forced labour since July 1, 2020. Enforcement has been close to nominal: two seizures in six years, a textile shipment in 2024 and frozen seafood in 2025, both from China. Bill C-35, introduced June 12, 2026 and still at first reading, would restructure the regime around a ministerial list of suspect goods, prescribed information requirements for listed goods, and deemed prohibition where those requirements go unmet. That is a functional move toward the U.S. model without adopting the U.S. presumption outright. Global Affairs Canada opened consultations on the regulatory approach on July 27.
Mexico. The Ministry of Economy regulation published February 17, 2023 and effective May 18, 2023 implements Mexico’s USMCA commitment. The proceeding runs through the Ministry of Labor and Social Welfare rather than the customs authority and can be triggered by third-party petition. Importers are expected to hold documentation showing their goods are not captured by lists the labor ministry publishes.
Four jurisdictions, four answers to who must prove what, to what standard, before which agency. A rebuttal package engineered for CBP’s clear and convincing standard is overbuilt for Brussels and misaddressed in Mexico City. A supplier attestation sufficient for a Canadian information request proves nothing under UFLPA.
// The tier measures the government, not the supply base
The most useful thing counsel can tell a sourcing team right now is that the Section 301 tier is not a risk rating.
Jordan is the cleanest illustration. On June 23, 2026, CBP issued two Withhold Release Orders under 19 U.S.C. § 1307 against garments produced in Jordan by Needle Craft Ltd. and Casual Wear Apparel L.L.C., citing ILO forced labor indicators including withheld wages, retention of identity documents, excessive overtime, and abusive conditions. One month later, Jordan landed in the 10 percent tier on the strength of its ART commitment and received conditional product exemptions designed to encourage follow-through.
Both things are true at once. The tariff tier reflects what a government enacted or promised. A WRO reflects what CBP found inside a factory. They measure different objects, and a buyer reading the 10 percent list as a safe-sourcing list has confused the two. That same month CBP issued a WRO against copper products manufactured in Serbia, and published importer guidance explaining the differences among the various forced labor authorities and their enforcement processes. The agency plainly expects confusion on this point.
// The risk that compounds
Divergence across regimes is manageable. Interaction is not.
Forced labor determinations generate records, and records do not respect jurisdictional boundaries. A CBP detention creates an administrative history an EU competent authority could reasonably treat as the seed of a substantiated concern. A Mexican labor ministry listing hands a petitioner in another market a ready-made referral. Bill C-35’s list of suspect goods, if enacted, becomes a second public register for NGOs, plaintiffs, and competing authorities to mine.
Companies have spent four years building programs around one question: can we rebut CBP. The harder question is whether an adverse outcome anywhere in the network becomes admissible, persuasive, or merely quotable everywhere else.
Containment strategies built for a single enforcer degrade when enforcement is networked.
// Exemptions with a foreign legislative calendar attached
One exemption category deserves scrutiny it has not received. USTR exempted certain products of Argentina, Bangladesh, Cambodia, Ecuador, El Salvador, the European Union, Guatemala, Indonesia, Jordan, Malaysia, Switzerland, Taiwan, and the United Kingdom on the stated basis that exemption would encourage those economies to fulfill commitments regarding forced labor import prohibitions, or to enact and effectively enforce one.
Read that as written. Those exclusions are instruments of persuasion, and their premise is a foreign government’s performance. An importer building landed cost on one is not relying on a classification position or an origin determination. It is relying on a parliament it does not lobby, on a timetable it cannot see. That dependency is invisible on the face of the HTS and belongs in the duty forecast as its own line of risk.
The reciprocal exposure is equally live. Ambassador Greer praised partners who moved quickly and said he looked forward to ensuring effective enforcement. The 10 percent rate rewards adoption. Nothing in the record makes it permanent if enforcement never arrives. Canada has held a prohibition for six years, seized two shipments under it, and sits in the 10 percent tier today. The September 15 training is how USTR builds the record distinguishing a statute on paper from a functioning border program, and that record cuts both ways.
// What to do before December 2027
Three things are worth doing before the EU regulation’s application date turns theoretical exposure into docketed proceedings.
- Inventory which jurisdictions can actually reach your imports, and map the burden of proof in each. The output is not a policy memo. It is a matrix telling the team, for a given shipment, who must prove what, to whom, and to what standard.
- Re-engineer the evidence package for portability. Documentation built to survive a UFLPA applicability review carries admissions and silences that read very differently to an authority operating under a cooperation duty rather than a presumption. Assume every package is eventually read by a regulator it was not written for.
- Price the conditional exemptions and the unestablished TRQs separately. A duty rate contingent on a foreign legislature, or on a quota USTR has not published, is a different category of risk from a classification position, and a blended rate hides it from the people who need to see it.
USTR’s framing is that the United States has been the only country to adopt and effectively enforce a forced labor import ban. Whether or not that was ever entirely right, it is a claim the agency is now working to make obsolete. Companies treating the September 15 convening as a press release will learn the difference in a venue of someone else’s choosing.