TRADE

China's Xinjiang Forced Labour System Is Inseparable from Its Exports. The U.S. Presumes It's All Tainted. Canada Presumes It's All Clean.

Canada's forced labour enforcement gap is now a CUSMA liability — and the July review deadline is weeks away.

Published June 3, 2026 · 30 sources · 14 min read

An empty customs inspection hall, sterile and brightly lit, with shipping containers visible through large windows, none of them opened

THE RECEIPTS Trade — The Gap

3.62×

Presumed Clean

Xinjiang's state-imposed forced labour system, the world's largest, is embedded across 17+ export industries. The region's exports grew 3.62× between 2021 and 2025. The U.S. built a rebuttable presumption blocking all goods unless proven clean. Canada uses a case-by-case system that has produced, at most, a handful of enforcement actions in five years. The government can't agree on the number.

XINJIANG EXPORTS (2025) 4,607.3亿元 (~US$63B)

EXPORT GROWTH (2021–2025) 3.62×

U.S. DETENTIONS (UFLPA) 16,755 SHIPMENTS

CANADA ENFORCEMENT (5 YRS) CONTRADICTORY RECORD

CUSMA REVIEW JULY 1, 2026

That's the receipt.

Full analysis · 30 sources · 14 min

Full analysis

Somewhere between a cotton field in southern Xinjiang and a shipping container arriving at a Canadian port, a question disappears. The question is simple: was this product made with forced labour? The United States decided three years ago that the only honest answer, for an entire region, was to assume yes and force importers to prove otherwise. Canada decided to ask the question one shipment at a time, with the burden on the government to prove the answer is yes. The result is an enforcement architecture that is structurally incapable of finding what it isn't designed to look for.

That gap was a policy curiosity until this morning. Now it's a tariff liability. The U.S. just proposed a 10% additional tariff on Canada, citing inadequate forced labour enforcement, with the CUSMA mandatory review deadline four weeks away. Canada's response: promise new legislation. The pattern should be familiar by now.

📊 How to read this article ▾

Green boxes contain documented facts with source citations you can verify.

Gold boxes contain our interpretation — clearly labeled analysis you can agree or disagree with.

Grey boxes provide context that both sides of the debate tend to leave out.

Red boxes list specific conditions that would change our assessment — our built-in accountability.

The narrative between the boxes is designed to be read on its own. The boxes add depth if you want it.

3.62×

Xinjiang export growth, 2021–2025

17+

Industries linked to forced labour

16,755

Shipments detained by U.S. CBP under UFLPA

Comparable Canadian enforcement actions

Key Facts — Documented

Xinjiang's reported exports in 2025 were 4,607.3亿元 (approximately US$63 billion), 3.62 times the 2021 level of 1,272.8亿元, according to Urumqi Customs annual briefs. (Note: absolute values pending unit confirmation from original Chinese-language customs releases; the 3.62× growth ratio is unaffected.) [9]
The Poverty Alleviation Through Labour Transfer programme is Xinjiang's primary coercive labour system, with labour transfers occurring more than 3 million times in 2022 alone, per the U.S. Department of Labor. [4]
The UFLPA establishes a rebuttable presumption that all goods from Xinjiang are produced with forced labour. Since implementation in June 2022, U.S. CBP has detained 16,755 shipments worth approximately US$3.7 billion. [1] [2]
Canada's forced labour import prohibition under Customs Tariff item 9897.00.00 has been in effect since July 2020. CBSA uses a case-by-case assessment process requiring the agency to demonstrate the forced labour nexus. [10] [12]
In Kilgour v. Canada, the Federal Court upheld CBSA's position that it lacks legal authority to apply a blanket regional presumption. [13]
Five UN Special Rapporteurs stated in January 2026 that Xinjiang's forced labour system "may amount to forcible transfer and/or enslavement as a crime against humanity." [8]
The USTR proposed a 10% additional tariff on Canada on June 2, 2026, under Section 301, citing inadequate forced labour enforcement. The investigation timeline is compressed to approximately five months, coinciding with the CUSMA review deadline of July 1, 2026. [5] [7] [27]

1. The System

In 2017, the Xinjiang Uyghur Autonomous Region launched a campaign of mass internment targeting Uyghurs and other predominantly Turkic Muslim groups. At its peak, an estimated one million or more people were detained in what Beijing called "vocational education and training centres." [8] The camps drew global condemnation, sanctions, and import bans. What drew less attention was the other system that expanded alongside them.

The Poverty Alleviation Through Labour Transfer programme predates the camps. It evolved from rural-urban migration policies in the early 2000s, and by the time international attention focused on the detention facilities, the transfer system had become the larger forced labour mechanism. The U.S. Department of Labor has described it as Xinjiang's "primary coercive labour system," recording more than 3 million transfer instances in 2022 alone. [4] (The DOL's figures draw on research by Adrian Zenz at the Victims of Communism Memorial Foundation, the leading but single-source academic researcher on Xinjiang labour transfers. [21]) In January 2026, five UN Special Rapporteurs concluded the coercion was severe enough that it "may amount to forcible transfer and/or enslavement as a crime against humanity." [8]

The programme works through local government work teams that mobilize ethnic minority labourers, transfer them in supervised groups, and monitor them on-site with political indoctrination sessions and surveillance. Research based on Chinese government documents found that workers who refused transfers were detained in re-education camps. [26] From 2021, Xinjiang's own policy documents show the programme intensified further, with a new "Unemployment Monitoring and Early Warning" system that effectively prevents workers from leaving their assigned jobs. [21]

The sectors touched by this system are not confined to the cotton fields that first drew scrutiny. The DHS Forced Labor Enforcement Task Force designates cotton, apparel, polysilicon, PVC, seafood, aluminum, and tomatoes as high-priority sectors, and in 2025 added caustic soda, copper, lithium, red dates, and steel. [3] The Coalition to End Uyghur Forced Labour has identified links across more than 17 global industries. [30] A joint investigation by the Bureau of Investigative Journalism, the New York Times, and Der Spiegel traced supply chains from Xinjiang labour transfers to factories producing washing machines, precision lenses, shoes, and processed food in multiple Chinese provinces. [36]

The point is not that every product leaving Xinjiang involves coerced labour. The point is that the system is so embedded in the region's economic architecture that separating clean goods from tainted ones requires granular, shipment-level tracing that the region's own governance structure is designed to obscure.

2. The Scale

Xinjiang's export economy did not contract under international pressure. It expanded, substantially.

According to Urumqi Customs annual briefs, Xinjiang's reported exports rose from 1,272.8亿元 in 2021 to 4,607.3亿元 in 2025, a factor of 3.62. (Chinese customs data uses 亿元 units, where 1亿 = 100 million RMB. The 2025 figure is approximately US$63 billion.) [9] Not all of this growth is attributable to forced labour sectors. Xinjiang sits at the crossroads of China's Belt and Road overland trade routes, and a significant portion of the region's trade figures reflect goods transiting through the region to Central Asian markets. But the growth in documented forced labour sectors is not incidental to this expansion. Xinjiang produces approximately 85% of China's cotton and more than 20% of the global supply. [25] China exports over 90% of the world's polysilicon, and 45% of all polysilicon originates from the Uyghur Region. [24]

In the solar industry, the response to U.S. enforcement was not to clean the supply chain but to bifurcate it. Companies created Xinjiang-free production lines for the American market while continuing to supply other global buyers from the same regional facilities. Because the U.S. market accounts for only 7% to 14% of Chinese solar companies' total production, researchers at Sheffield Hallam University found that "the vast majority of modules produced globally continues to have exposure to the Uyghur Region." [29] The raw production volume of Xinjiang polysilicon has actually increased alongside global demand, even as the region's global market share declined.

So what does this mean for trade enforcement? It means the contamination is structural, not episodic. A country that wants to keep forced labour goods out of its market cannot do it by inspecting individual shipments. The supply chain is designed to make that impossible.

3. The Wall

The United States concluded exactly that. The Uyghur Forced Labor Prevention Act, signed in December 2021 and implemented in June 2022, created a rebuttable presumption: all goods mined, produced, or manufactured wholly or in part in the Xinjiang Uyghur Autonomous Region are presumed to be made with forced labour. [1] The burden flips. It is not the government's job to prove the goods are tainted. It is the importer's job to prove they are clean.

The enforcement record reflects the design. By August 2025, U.S. Customs and Border Protection had detained 16,755 shipments valued at nearly US$3.7 billion under the UFLPA, according to CSIS analysis of the CBP dashboard. [23] In the first year alone, more than US$1.3 billion in products were detained and at least 679 shipments were denied entry outright. [2]

The system is not without contradictions. CBP detentions dropped sharply in 2025, falling to just 14 shipments in July of that year, an unprecedented low. [23] In total, the value of goods detained under UFLPA in 2025 fell sharply from the prior year. Reported figures vary by cut-off date and methodology: the Globe and Mail reported a decline to US$171 million from US$1.38 billion; Reuters, citing CBP data in February 2026, reported US$187.7 million down from US$1.78 billion. [31] Whether this reflects successful deterrence, supply chain bifurcation, or a policy choice by the Trump administration to deprioritize enforcement while stabilizing commercial relations with Beijing is contested. What is not contested is that the legal architecture exists, the presumption is operational, and it produced measurable enforcement results for three years.

4. The Absence

Canada's approach is the structural opposite.

Customs Tariff item 9897.00.00, effective since July 2020, prohibits the import of goods mined, manufactured, or produced wholly or in part by forced labour. [10] The prohibition was adopted to meet Canada's CUSMA commitments. [14] Bill S-211, effective January 2024, expanded the prohibition to include child labour and introduced annual supply chain reporting requirements for large companies. [11] On paper, the legal tools exist.

In practice, the enforcement architecture inverts the American model. Under CBSA Memorandum D9-1-6, the burden of proof sits with the government. CBSA must establish, on a case-by-case basis, that a specific shipment involves forced labour before it can be detained. [12] There is no regional presumption, no entity list, no sector-wide enforcement priority. Each container arriving at a Canadian port is, by default, presumed clean.

Human rights advocates challenged this framework in court. In Kilgour v. Canada, they argued CBSA should presumptively prohibit imports from Xinjiang, mirroring the American approach. CBSA responded that it lacked the legal authority for a blanket regional presumption. The Federal Court agreed. [13]

The enforcement record reflects the architecture. In June 2024, a parliamentary committee intervention stated Canada had not seized a single shipment under the forced labour prohibition. By November 2025, a separate committee intervention cited 50 interceptions and one denial of entry. Globe and Mail reporting referenced two blocked shipments. [31] [32] These figures are not reconcilable from public sources. No authoritative CBSA statistical output providing a definitive cumulative count appears to exist in the public record. The government's own accounting of its enforcement record is, itself, contradictory.

Meanwhile, Chinese customs data as reported by the Globe and Mail shows Xinjiang exports to Canada jumped 160% year-over-year to US$601 million in 2025. [31] These are Chinese-reported figures for goods shipped, not goods accepted. But the trajectory is not ambiguous.

In October 2024, Global Affairs Canada launched public consultations on strengthening enforcement of the forced labour import ban, acknowledging the existing framework's limitations. [15] As of June 2026, no new enforcement mechanism has been implemented.

5. The Delegation

In January 2026, Prime Minister Mark Carney travelled to Beijing and negotiated an agreement to import 49,000 Chinese-manufactured electric vehicles into Canada at reduced tariff rates. [18] A Privy Council Office report subsequently submitted to Parliament stated that "human rights and foreign interference were not brought up proactively" during Carney's meeting with Xi Jinping. The Prime Minister's Office later described the document as "submitted in error" and sent a corrected version to Parliament. [17]

In March 2026, the House of Commons Standing Committee on Industry and Technology held hearings on the EV agreement's forced labour implications. Margaret McCuaig-Johnston, a senior fellow at the University of Ottawa's Institute for Science, Society and Policy, testified that aluminum, polysilicon, and critical minerals essential for EV batteries and components are tied to Xinjiang's coercive labour transfer programmes. [16]

Liberal MP Michael Ma, part of Carney's official delegation to Beijing, questioned the expert's evidence. He asked whether she had personally observed forced labour with her own eyes. The exchange drew condemnation from Uyghur advocacy organizations and Conservative members of the committee. Ma later issued a public apology. [16] [33] [37]

The incident crystallized something that the enforcement data already showed. The question is not whether Canada's political leadership is aware of the forced labour system. The question is whether the country's enforcement architecture, trade posture, and diplomatic practice are designed to act on that awareness. The documented record suggests they are not.

6. The Deadline

On June 2, 2026, the United States Trade Representative released the findings of a Section 301 investigation into forced labour enforcement across 60 economies, proposing a 10% additional tariff on countries found to have inadequate enforcement regimes. Canada is on the list. Reuters reported that goods qualifying under CUSMA/USMCA rules of origin would be exempt from the proposed tariff, though the scope of that exemption remains subject to public comment. [34] [5] [6] [7]

The investigation was launched in March 2026, shortly after the U.S. Supreme Court struck down the "Liberation Day" tariffs under Section 122. Legal analysts at Norton Rose Fulbright noted the Section 301 forced labour track provides an alternative legal vehicle for tariffs the courts removed, with a compressed five-month timeline that coincides with the CUSMA mandatory review deadline of July 1, 2026. [27] The USTR's 2026 National Trade Estimate Report specifically alleged that Canada had failed to prevent goods made with forced labour from entering the U.S. market. [6]

Carney's response came the same day the tariff proposal was released. The Prime Minister said the tariffs were "not a surprise" given the U.S. had signalled its intent to replace the Section 122 tariffs, and announced that the government would "soon introduce legislation" on forced labour in supply chains. [19] Conservative finance critic Adam Chambers said the current laws "are not working in the way that they should." [20]

CUSMA's labour chapter already requires Canada to prohibit forced labour imports and to cooperate with the United States on enforcement. [14] The agreement's mandatory review is four weeks away. The U.S. has now formally documented its position that Canada is not meeting its obligations. Whether the forced labour tariffs survive legal challenge is uncertain, but the negotiating leverage they create for the CUSMA review is not.

Plain English: China's Xinjiang region runs the world's largest state-imposed forced labour system. It touches cotton, solar panels, aluminum, car parts, seafood, tomatoes, and at least a dozen other industries. The U.S. assumes everything from the region is tainted and makes importers prove otherwise. Canada assumes everything is clean and requires the government to prove otherwise, one shipment at a time. The result: the U.S. has detained over 16,000 shipments. Canada's enforcement record is so thin the government contradicts itself on the numbers. Now Washington is using that gap to justify new tariffs, just weeks before Canada's trade agreement comes up for review.

Context — What Both Sides Omit

Critics of Canada's enforcement tend to omit two things. First, the U.S. enforcement record itself collapsed in 2025: CBP detentions dropped to US$171 million from US$1.38 billion the year before, raising questions about whether the American model is being sustained under the current administration. Second, the Section 301 forced labour tariffs are not purely a human rights instrument. They emerged as a legal replacement for tariffs the Supreme Court struck down, which means the enforcement critique, however legitimate, is also serving as a vehicle for protectionist trade policy.

Defenders of the Canadian approach tend to omit the structural consequence of the case-by-case model. If the burden is on CBSA to prove each shipment's forced labour nexus, and CBSA lacks the intelligence infrastructure, staffing, and legal authority to conduct those investigations at scale, then the system is not a different enforcement philosophy. It is the absence of enforcement with a legal framework around it. The Federal Court ruling in Kilgour confirmed that this is not a resource problem the government can fix administratively. It requires legislation.

Interpretation

The structural finding here is not that Canada tolerates forced labour. It is that Canada built an enforcement architecture that is incapable of producing a different outcome. The case-by-case model, the burden on government, the absence of a regional presumption, the Federal Court ruling foreclosing administrative workarounds: these are not individual policy failures. Together, they constitute a system that presumes goods are clean by default and lacks the tools to challenge that presumption at scale. The forced labour import ban exists on the statute books. Its enforcement does not exist in practice at any meaningful volume.

This matters beyond the moral dimension because it is now a trade liability. The United States has formally documented Canada's enforcement gap as a basis for tariffs, weeks before the CUSMA mandatory review. The Prime Minister's response, announcing forthcoming legislation on the same day the tariff was proposed, follows a pattern this publication has documented before: announce the intention to legislate in response to a structural failure, without addressing the enforcement mechanism that made the existing law toothless.

Counter-interpretation: A thoughtful defence of Canada's position would argue that the rebuttable presumption model is a blunt instrument with significant collateral damage. It captures goods with no forced labour nexus, imposes compliance costs that fall disproportionately on small importers, and gives the U.S. government unilateral power to restrict trade under a human rights banner that can be selectively applied. Canada's case-by-case model, in this view, is not weakness but precision. As trade lawyer William Pellerin has characterized the distinction, it is a scalpel rather than a sledgehammer. [32] The premise would be that enforcement resources should be concentrated on high-risk sectors and specific entities rather than applied regionally. This argument has structural merit. Its weakness is empirical: the scalpel has, by any public measure, cut almost nothing.

What Would Change This Assessment

If CBSA published enforcement data showing 100 or more forced-labour shipment detentions, or multiple final exclusions across high-risk sectors, since July 2020, the article's near-zero enforcement characterization would require substantial revision.
If Canada's promised legislation introduces a rebuttable presumption or equivalent sector-wide enforcement mechanism, and that mechanism produces measurable enforcement actions within its first year of operation, the characterization of Canada's approach as structurally incapable would weaken. (Legislation announced but not yet tabled does not satisfy this condition.)
If the Section 301 forced labour tariffs are struck down by U.S. courts or withdrawn before the CUSMA review, the trade liability framing would lose its immediate urgency. The enforcement gap itself, however, is documented independently of the tariff dispute.
If U.S. CBP enforcement volumes return to 2023–2024 levels in 2026, the critique of American enforcement inconsistency in the Context box would weaken.
If the original Urumqi Customs releases, when independently reviewed, show a materially different growth trajectory than the 3.62× ratio reported here, the export-growth dimension of the article would require correction.

Sources (30)

Uyghur Forced Labor Prevention Act (UFLPA), Pub. L. 117-78, signed December 23, 2021, implemented June 21, 2022. Establishes rebuttable presumption for goods from Xinjiang. Congress.gov
U.S. Customs and Border Protection, UFLPA Statistics Dashboard. Enforcement totals for shipments detained, reviewed, denied, and released since June 2022. CBP.gov
DHS Forced Labor Enforcement Task Force (FLETF), 2025 Strategy Update, August 19, 2025. Designates high-priority enforcement sectors. DHS.gov
U.S. Department of Labor, Bureau of International Labor Affairs (ILAB), "Against Their Will: The Situation in Xinjiang." Documents Poverty Alleviation Through Labour Transfer as Xinjiang's primary coercive labour system; 3 million+ transfer instances in 2022. DOL.gov
USTR Section 301 investigation announcement, March 12, 2026. Investigations into 60 economies under Section 301(b) of the Trade Act of 1974. USTR.gov
USTR, 2026 National Trade Estimate Report on Foreign Trade Barriers. Alleges Canada failed to prevent forced labour goods from entering U.S. market. USTR.gov
USTR report, released June 2, 2026. Proposes 10% additional tariff on Canada and others for inadequate forced labour enforcement under Section 301. USTR.gov
UN Office of the High Commissioner for Human Rights, press release, January 22, 2026. Five Special Rapporteurs statement: Xinjiang forced labour system "may amount to forcible transfer and/or enslavement as a crime against humanity." Xinjiang five-year plan (2021–2025) projects 13.75 million labour transfer instances. OHCHR.org
Urumqi Customs / General Administration of Customs, Xinjiang annual customs briefs. Verified export figures (in 亿元): 2021: 1,272.8; 2022: 2,091.2; 2024: 3,699.8; 2025: 4,607.3. Unit confirmation from original Chinese-language releases pending. Primary customs brief documents for 2021, 2024, and 2025; customs-hosted summaries for 2022. Urumqi Customs
Canada Customs Tariff, tariff item 9897.00.00. Prohibits import of goods mined, manufactured, or produced wholly or in part by forced labour. Effective July 2020. CBSA
Fighting Against Forced Labour and Child Labour in Supply Chains Act (Bill S-211), S.C. 2023, c. 9. Effective January 1, 2024. Expanded prohibition to child labour; requires annual corporate supply chain reporting. Justice.gc.ca
CBSA Memorandum D9-1-6: Prohibition on importation of goods produced by forced labour. Case-by-case enforcement framework with ESDC support. CBSA
Kilgour v. Canada, Federal Court. Upheld CBSA's position that it lacks legal authority to apply a blanket regional presumption; enforcement limited to case-by-case assessment of specific shipments.
Canada-United States-Mexico Agreement (CUSMA), Chapter 23, Labour. Requires parties to prohibit forced labour imports and cooperate on enforcement. Global Affairs Canada
Global Affairs Canada announcement, October 16, 2024. Launch of public consultations on strengthening enforcement of the forced labour import ban. Canada.ca
House of Commons Standing Committee on Industry and Technology, testimony, March 26, 2026. Margaret McCuaig-Johnston testimony on forced labour links to EV supply chains; Michael Ma questioning. OurCommons.ca
Privy Council Office report to Parliament, March 2026. States "human rights and foreign interference were not brought up proactively" by PM Carney during January 2026 meeting with Xi Jinping. PM's Office subsequently described document as "submitted in error"; corrected version sent to Parliament. PCO
Canada-China EV agreement, January 2026. Import of 49,000 Chinese-manufactured EVs at reduced tariff rates, negotiated during PM Carney's Beijing visit.
PM Mark Carney, statement, June 3, 2026, Ottawa. Tariffs were "not a surprise"; government will "soon introduce legislation" on forced labour in supply chains. [reported: 35]
Adam Chambers, Conservative finance critic, statement, June 3, 2026. Current laws "are not working in the way that they should." [reported: 35]
Adrian Zenz / Victims of Communism Memorial Foundation. Research finding: Xinjiang labour transfers reached new heights, with the "Unemployment Monitoring and Early Warning" system preventing workers from leaving assigned jobs. Note: single-researcher dependency. VoC Memorial Foundation
Adrian Zenz, RealClearWire op-ed, May 15, 2026. Claims Xinjiang exports to Canada/UK/EU rose 465% (2021–2025). Note: underlying Xinjiang-by-destination dataset not independently verified. Not used as a finding in this article. RealClearWire
CSIS, "Assessing the Impact of the Uyghur Forced Labor Prevention Act After Three Years," August 29, 2025. Reports 16,755 detained shipments (~US$3.7B); sharp April–May 2025 enforcement downturn; 14 shipments in July 2025. CSIS.org
CSIS, "The Uyghur Forced Labor Prevention Act Goes into Effect." China exports 90%+ of world polysilicon; 45% originates from XUAR. CSIS.org
CSIS, "The WTO Can Help Shine a Spotlight on Forced-Labor Practices in Xinjiang's Cotton Industry." Xinjiang produces ~85% of China's cotton, 20%+ of global share. CSIS.org
New Lines Institute, "Coercive Labor in Xinjiang" (Zenz). "Strong evidence that the production of the majority of Xinjiang's cotton involves a coercive, state-run program targeting ethnic minority groups." New Lines Institute
Norton Rose Fulbright, legal analysis, June 2, 2026. Section 301 timeline compressed to ~5 months, coinciding with CUSMA review deadline July 1, 2026. CUSMA-compliant goods exemptions may not apply to Section 301 tariffs.
Human Rights Watch, "Asleep at the Wheel," February 2024. Documents forced labour links in automotive sector: aluminum, polysilicon, critical minerals in EV batteries and components. HRW.org
Sheffield Hallam University / Laura Murphy research. U.S.-bound supply chains: 7–14% of Chinese solar companies' total production; "vast majority of modules produced globally continues to have exposure to the Uyghur Region." Sheffield Hallam University
Coalition to End Uyghur Forced Labour. More than 17 global industries linked to state-sponsored Uyghur forced labour. EndUyghurForcedLabour.org

Reporting sources (not counted in the 30-source total; used for direct quotes from named officials and factual context):

Globe and Mail, April 7, 2026. Chinese customs data showing Xinjiang exports to Canada rose 160% YoY to US$601M in 2025; CBP detained US$171M in 2025, down from US$1.38B. Globe and Mail
Globe and Mail, March 27, 2026. Pellerin "scalpel vs. sledgehammer" characterization; enforcement gap reporting.
CBC News, March 26–27, 2026. Michael Ma committee incident; apology; background.
CBC News, June 3, 2026. Carney "not a surprise" response; CUSMA forced labour obligations context.
Canadian Press, June 3, 2026. Carney and Chambers statements on new forced labour legislation.
Bureau of Investigative Journalism / New York Times / Der Spiegel, joint investigation, 2025. Traced Xinjiang labour transfer supply chains to factories producing consumer goods across multiple Chinese provinces.
Campaign for Uyghurs / Uyghur Rights Advocacy Project / World Uyghur Congress, press releases, March 2026. Condemning Ma remarks; documenting forced labour evidence base.

No corrections have been issued for this article. See our corrections policy.

This article was produced using AI-assisted adversarial verification. Read how and why.