Introduction: the market without a map
The international regulation of trade has long tolerated some untidiness. For many years GATT and a web of preferential schemes coexisted comfortably, and Article XXIV of GATT was designed on the assumption that regionalism would remain an exception within a multilateral system with a coherent core. It is this assumption that is now missing in digital trade. There is no WTO equivalent of the GATT tariff schedules, no modal classification akin to that of the GATS, and no text on data flows, data localisation or source code agreed by all WTO members. Instead, there is a mosaic of treaties: Chapter 19 of the United States-Mexico-Canada Agreement, which prohibits data localisation requirements and protects source code in close-to-absolute terms;1 the Digital Economy Partnership Agreement between Chile, New Zealand and Singapore, whose sixteen modules are designed as building blocks that other economies may join or adopt in whole or in part;2 and the India-UAE Comprehensive Economic Partnership Agreement, which includes a chapter on digital trade but expressly excludes it from the Agreement’s dispute settlement mechanism, its commitments being framed largely as undertakings to “endeavour”.3 Each of these treaties answers the fundamental questions of digital trade governance in its own way, and none of them was drawn up against a baseline set of rules established by the WTO.
The core of this paper is the assertion that the growing body of conflicting digital trade regimes is not the disease itself but a symptom of an underlying defect, namely the inability of the World Trade Organization, after the 1998 Declaration on Global Electronic Commerce brought digital trade into focus,4 to create a multilateral framework that could serve as the basis for later bilateral agreements. The Joint Statement Initiative on E-Commerce set out to do so in 2017,5 and in December 2024 seventy-two co-sponsors circulated the concluded text of the Agreement on Electronic Commerce.6 Yet at the last General Council meeting of 2025 members still could not agree to incorporate the plurilateral agreement into the WTO legal system under Annex 4 of the Marrakesh Agreement, seven members expressing concerns; at the February 2025 meeting India and Indonesia had objected, while South Africa, Pakistan and Bangladesh had said they were not yet ready to agree, and the interim arrangements that a group of members adopted at the Fourteenth Ministerial Conference in March 2026 operate pending, not instead of, that incorporation.7 As a result, the WTO’s own initiative on digital trade is caught up in the very fragmentation it was supposed to remedy: a text exists, but it has no effect as WTO law, and countries continue to pursue bilateral deals in its shadow.
This paper makes its argument in three stages. Section II outlines the structural deficiency at the heart of the multilateral system. Sections III to V analyse the three gaps that result from it (regulatory arbitrage, the compliance burden on small businesses, and developmental exclusion) as symptoms of the same disease rather than independent phenomena. Section VI tests the diagnosis on two real-world instruments, DEPA and the India-UAE CEPA. Sections VII and VIII propose and examine modular and interoperable remedies.
The underlying issue: the multilateral vacuum at the heart of the web
Every downstream issue discussed in this paper stems from a common institutional failure: the WTO has never agreed on what a “baseline” digital trade obligation entails. The GATS was negotiated in 1994, when commercial internet trade did not yet exist as an identifiable practice, and its four modes of supply, its general exceptions in Article XIV and its scheduled services commitments could not possibly have answered questions such as whether the transmission of data constitutes a “service”, whether data localisation requirements are a valid domestic regulation or a trade restriction, or whether source code is a trade secret, a good or something sui generis.8 Although GATS Article XIV allows members to take measures “necessary to protect public morals” or “necessary to secure compliance” with domestic laws not inconsistent with the Agreement, including laws relating to the protection of privacy,9 the Appellate Body’s application of the necessity test in US – Gambling and in China – Publications and Audiovisual Products (the latter under the parallel GATT Article XX(a)) shows how unpredictable that exception is.10 In China – Publications, in particular, the Panel and the Appellate Body proceeded on the basis that China had chosen a high level of protection of public morals, yet the Appellate Body upheld the finding that China had not demonstrated its measures to be “necessary”, because China had not shown that a less trade-restrictive alternative proposed by the United States was not reasonably available.11 Transplanted to disputes over data localisation or source-code disclosure, this precedent leaves room for fact-specific assessment rather than providing a precise legal standard, precisely because the GATS was never amended to give guidance on digital trade obligations.
Multilateral attempts to fill this vacuum have failed for want of consensus, while plurilateral negotiations have not achieved multilateral status. The Joint Statement Initiative on E-Commerce, launched in December 2017 by seventy-two WTO members at the Buenos Aires Ministerial Conference, began as exploratory work towards future negotiations, and when members confirmed in January 2019 their intention to negotiate, they did so seeking an outcome “with the participation of as many WTO members as possible”,12 thus implicitly admitting that multilateral consensus would be hard to reach. Seven years on, with ninety-one participants, the co-convenors announced a stabilised text of the Agreement on Electronic Commerce in July 2024,13 although in October 2023 the United States had withdrawn its support for the core data-flow, data-localisation and source-code provisions, citing, in Henry Gao’s account, the need to balance the right to regulate in the public interest, even though it had advocated such provisions in the USMCA.14 The moratorium on customs duties on electronic transmissions, first agreed in the 1998 Declaration on Global Electronic Commerce, illustrates the same process: a provisional commitment, renewed at successive Ministerial Conferences for more than a quarter of a century without ever being made permanent, lapsed on 30 March 2026 when the Fourteenth Ministerial Conference failed to extend it.15 The absence of a durable baseline is therefore not an accident but the result of a series of deliberate decisions by the membership.
It is this absence, and not the content of any one agreement, that is causing the confusion in digital trade governance. Because there is no WTO text that a country drafting an FTA chapter can use as a point of reference, every negotiator drafts on a blank page, and each fills it differently. This is what causes the specific problems discussed in the following three sections.
The first offshoot: regulatory arbitrage and the emergence of data havens
The most immediate effect of the lack of a multilateral baseline is regulatory arbitrage. USMCA Article 19.12 prohibits, without any exception of its own, any requirement that a covered person use or locate computing facilities in a Party’s territory as a condition for conducting business in that territory, subject only to the GATS Article XIV(a)-(c) exceptions incorporated by Article 32.1(2).16 Other regional trade agreements, and the unilateral domestic legislation of countries that are party to no regional digital trade agreement, allow discretionary localisation. Corporations therefore have a real choice of legal regime in which to set up their servers and through which to route their data. Michael Geist, discussing the USMCA, observes that Canadian negotiators presumably read the incorporated GATS privacy exception as an opening that would let British Columbia and Nova Scotia keep their data localisation laws, which suggests that even three parties using the same treaty text could reach very different outcomes.17 Multiply that difference across dozens of treaties (over 20 per cent of the 595 regional trade agreements in force in October 2023 included e-commerce provisions18) and the incentive to route data through whichever jurisdiction enforces its rules most weakly becomes structural. The phenomenon has been called the “spaghetti bowl” of digital trade, borrowing the term long applied to overlapping preferential trade agreements,19 or, in trade economists’ terminology, the “digital noodle bowl”.20
Empirical work supports this account. Evenett, Fritz and Giardini of the Global Trade Alert argue that domestic digital regulations need not be discriminatory to fragment the world’s digital economy: heterogeneity among non-discriminatory regulations alone can obstruct cross-border digital trade, and WTO rules and digital trade agreements “offer important directions but insufficiently mitigate heterogeneity”.21 Korgun’s study of South Korea’s place in the evolving digital trade architecture likewise treats regulatory fragmentation as a key feature of the current landscape, precisely because there is no common international standard on digital trade governance to which countries can align.22 The risk is a race that favours laxer regulation, in which countries compete not on the efficiency of their digital infrastructure but on the leniency of their data protection enforcement.
The second offshoot: the compliance maze and the squeeze on SMEs
Large multinational technology companies can arbitrage between competing systems; small and medium enterprises cannot. They lack the legal expertise needed to comply with conflicting obligations on cross-border data transfer, digital invoicing standards and consumer protection disclosure, obligations that vary with the FTA chapters to which their export markets subscribe. The digital spaghetti bowl has been discussed from precisely this angle: as the number of e-commerce provisions in regional trade agreements rose markedly after the USMCA took effect in 2020, these agreements wove “intricate webs of multiple and overlapping memberships” that impede rather than facilitate seamless digital trade, a burden felt most by smaller businesses.23 In the absence of a WTO minimum standard, a company selling in USMCA, DEPA and CEPA markets must keep up with three different compliance regimes on data localisation alone: the USMCA prohibits it outright; DEPA affirms a prohibition that, under its Annex I understanding, creates no rights or obligations between the Parties under DEPA itself; and the India-UAE CEPA does not ban it at all.24
The problem is compounded by the uneven enforceability of the three instruments’ digital trade commitments. USMCA Chapter 19 is subject to the Agreement’s general state-to-state dispute settlement under Chapter 31. DEPA has a dispute settlement module of its own, but it excludes the core data-flow and localisation articles from it, and Elms observed when the text was concluded that many of DEPA’s provisions remained tentative, with commitments for future work and consultation, and that most working elements of its dispute settlement system remained to be developed.25 The India-UAE CEPA excludes its digital trade chapter from dispute settlement altogether, thereby confirming, in one commentary’s words, “the aspirational nature of the chapter”.26 For an SME, the absence of an enforceable mechanism is no less problematic than the absence of a binding rule: there is no assurance that a competitor’s non-compliance in another market will ever be addressed.
The third offshoot: the digital divide and the room where LDCs are not
The third problem concerns distribution, not exchange. The rules of digital trade are mostly made by and for countries that already have digital systems, and least-developed countries are largely absent from the rooms where these rules are decided. When the e-commerce negotiations concluded in December 2024 they involved ninety-one participants, of which only five were least-developed countries: Benin, Burkina Faso, the Gambia, Lao PDR and Myanmar. None of the developing Pacific Island countries took part.27 UNCTAD’s Digital Economy Report shows what this gap represents. The United States and China account for 90 per cent of the market capitalisation value of the world’s seventy largest digital platforms, more than 75 per cent of the world market for public cloud computing and 75 per cent of all patents related to blockchain technologies. Meanwhile, only one in five people in the least-developed countries uses the internet, compared with four in five in developed countries.28
The gap is widening, not narrowing. In 2024 digitally deliverable services made up more than 60 per cent of total services exports in advanced economies, but only 15 per cent in the least-developed countries, and developing countries together accounted for barely a fifth of global digital services exports.29 This asymmetry matters for the legitimacy of the rules. It means that the “rules” filling the WTO’s gap are not neutral technical standards; they reflect the goals of the countries that are strong in digital trade. Developing countries that are party to no such agreement bear the costs of following rules made by others, such as the USMCA or DEPA, with no say in how those rules are made. In the words of the Group of 77 and China, many developing economies take part in the digital economy “mainly as guests on digital platforms”.30
A genuine WTO agreement, made by all members on the one-member-one-vote basis of the WTO, would at least give least-developed countries a place at the table where the rules that affect them are made. No group of selected countries, however open it tries to be, can do that.
Two experiments at the edge of the vacuum: DEPA and the India-UAE CEPA
The analysis above can be tested against two instruments that have tried, in different ways, to work without a WTO baseline.
Singapore, Chile and New Zealand signed the Digital Economy Partnership Agreement electronically in June 2020, deliberately designing it as a pathfinder agreement of sixteen modules. The modules cover data issues, digital identities, e-invoicing, artificial intelligence governance and small-business cooperation.31 Other economies may accede to the agreement or take up individual modules, in whole or in part, in other agreements or in domestic policy, rather than treating DEPA as a single indivisible package. The Republic of Korea substantially concluded its accession negotiations in June 2023 and became the first new Party in May 2024, Costa Rica’s accession followed in 2026, and China and Canada have both applied to join, evidence that the modular design has lowered the cost of joining relative to a comprehensive treaty.32 Yet DEPA’s own structure shows the limits of modularity without a binding baseline. Because DEPA’s core data commitments create no rights or obligations between the Parties and sit outside its dispute settlement module, and because many of its provisions are framed as commitments to future cooperation, modules can be set aside whenever a signatory’s political focus shifts elsewhere. China’s accession working group, established in August 2022, had still not concluded when Costa Rica’s accession was announced in September 2026, which suggests how difficult it is to align Beijing’s data-governance system with DEPA’s commitments.33 The lesson is that modularity helps countries reach the table but does not ensure that they keep their promises once they arrive, unless some structure of binding rules is added on top of the modules.
India’s approach under the CEPA signed with the United Arab Emirates in February 2022 sits at the other end of the ambition scale. It was India’s first free trade agreement to include a dedicated digital trade chapter, covering paperless trading, consumer protection, cross-border information flows and cooperation on e-payments and e-invoicing.34 It commits each Party to maintain its current practice of not imposing customs duties on electronic transmissions between them, a practice that either Party may adjust in light of further WTO decisions on the moratorium, a point of practical weight now that the multilateral moratorium has lapsed.35 But the chapter is an “endeavours” instrument: it is excluded from the CEPA’s dispute settlement process, sets no standard for the personal data protection framework it asks each party to endeavour to adopt, and was negotiated and signed before India enacted the Digital Personal Data Protection Act, 2023. Its cross-border data-flow language was thus left to operate against a domestic legal background that was then entirely missing.36 The DPDP Act that followed adopts a model that permits cross-border transfers by default but empowers the Central Government to restrict transfers to countries or territories it notifies, a “negative list” method; section 16, which contains that power, has not yet come into force, the commencement notification of 13 November 2025 having deferred it by eighteen months. This is a middle ground between, on the one hand, the mirroring of sensitive and the localisation of critical personal data proposed in the Personal Data Protection Bill, 2019 and the stricter localisation recommended by the Joint Parliamentary Committee in 2021, and, on the other, the liberal USMCA-style ban on localisation that India has never accepted.37 Commentary on the CEPA’s operation has already observed that the parties’ data protection laws leave cross-border data transfers for fintech and e-commerce businesses “in limbo”,38 and how the CEPA’s aim of free information flow will interact with any countries India notifies under section 16 remains unsettled.
Taken together, DEPA and the India-UAE CEPA show that the absence of a WTO baseline does not merely create different rules; it creates different kinds of instrument. One is voluntary and modular, the other framed in best-endeavour terms and excluded from dispute settlement. Each makes sense on its own terms, but the two cannot be compared, and neither can be turned into a general multilateral standard without major changes.
From knot to lattice: modularity and interoperability as solutions
The central problem is the absence of an agreed baseline, and DEPA shows that voluntary modularity alone cannot replace one. The better solution is to bring modularity into the WTO’s own process and to pair it with binding minimum commitments and formal interoperability bridges between existing regional regimes.
First, a modular WTO framework would let members join self-contained disciplines, such as e-invoicing standards or electronic authentication, without having to settle at the outset the contested questions of cross-border data flows or source-code disclosure.
This is broadly what the Joint Statement Initiative already attempts. The Agreement on Electronic Commerce stabilised in July 2024 contains no market access commitments, provides exceptions for prudential measures, personal data protection, security and Indigenous Peoples, and includes a mechanism to review the agreement periodically and to return to negotiations in the future.39
The real problem is not the design of the modules but their status. As Section II showed, even a text negotiated by ninety-one participants and co-sponsored by seventy-two members cannot be added to Annex 4 of the Marrakesh Agreement as a plurilateral agreement without consensus, and that same consensus rule has stalled negotiation from the start.40
Second, interoperability bridges are recognition arrangements: parties agree to accept each other’s data protection regimes, even where the texts differ, so long as each meets a baseline of security and privacy. Such bridges offer a path that does not require all WTO members to agree at the outset.
The European Union has concluded with Singapore its first standalone digital trade agreement, which entered into force on 1 February 2026, and China has applied to join DEPA even though doing so would create tension with DEPA’s commitments. Both are tests of interoperability, and both are ways for larger economies to connect with an existing framework instead of building a new one.41
Applied at the WTO, this idea would formally recognise DEPA, USMCA Chapter 19 and similar “high-ambition” regional agreements as WTO-consistent building blocks, much as GATT Article XXIV treats customs unions and free-trade areas as exceptions to most-favoured-nation treatment provided they meet its conditions, instead of treating each agreement as an experiment to be judged only later in dispute settlement under GATS Article XIV. Neither solution is free of cost. Both require WTO members to accept a lower starting level of ambition than a single comprehensive multilateral text would offer. The alternative, letting regional and bilateral rules multiply without any reference point, has already produced arbitrage, a compliance burden for small and medium enterprises and developmental exclusion, the problems set out in Sections III to V. They will not disappear if nothing is done.
What the lattice costs: economic, legal and geopolitical implications
Adopting these frameworks, even as a second-best alternative to a full WTO baseline, has consequences of three different kinds.
Economically, reducing compliance friction should expand the volume of cross-border trade and benefit small and medium enterprises disproportionately. Supporters of the Agreement on Electronic Commerce say that it creates a more predictable digital trading framework for businesses, workers and consumers, even though it contains no market access commitments.42 The same effect is stronger where interoperability bridges reduce the number of compliance regimes a company must follow.
Legally and in regulatory terms, however, interoperability requires regulators to give up some of what they may regard as sovereignty over data governance in exchange for international recognition of equivalence. This trade-off is politically costly in countries such as India, which have historically resisted disciplines prohibiting localisation, a tool its advocates defend as aiding law-enforcement access and domestic industry development.43 The DPDP Act’s negative-list compromise shows how hard the trade-off is even at the domestic level, before any international mutual-recognition requirement is added. A WTO-level interoperability mechanism would need a domestic-regulation exception modelled on the general exceptions of GATS Article XIV and GATT Article XX, as the Appellate Body applied them in US – Gambling and China – Publications, which let members define their objectives and level of protection for themselves while still testing whether a less trade-restrictive alternative is reasonably available.44
Geopolitically, the main risk is that “interoperability” becomes a byword for whichever major power sets its domestic baseline first. The United States, the European Union and China each start from a different position. Under the USMCA model the United States essentially prohibits data localisation and source-code disclosure requirements; under the GDPR the European Union follows a rights-based model; and China follows a security-and-sovereignty model that requires the localisation of certain categories of data. Each wants its own model to become the standard against which other members are judged “equivalent”.45 That the United States withdrew its support for the JSI’s core data-flow rules in October 2023 while those rules remain in the USMCA suggests that even the largest digital powers are reluctant to bind themselves multilaterally to standards they are able to enforce bilaterally against smaller partners.46
Conclusion: strategic lessons for the WTO
The confusion in digital trade governance is not caused by the number of trade agreements. It comes from the lack of a strong multilateral foundation. Agreements such as the USMCA, DEPA and the India-UAE CEPA are not poorly written; each reflects the needs of the countries that joined it. But no single agreement, nor even all three together, can replace what a true multilateral system provides.
That system must come from the World Trade Organization. Only the WTO can create a baseline, a standard that every member must follow. This baseline would set a floor for digital trade law below which no country’s domestic rules could fall. It would include a real dispute settlement mechanism with binding force. And it would ensure that developing and least-developed countries have full representation at the table, so that they are not left behind.
Three clear lessons follow from this analysis. First, the WTO should turn its single-tier Agreement on Electronic Commerce into something more flexible: it should make it modular, like DEPA, but with binding minimum commitments on key issues. These core issues should include a customs duty moratorium, non-discrimination for digital products and basic protection of personal data, and these core elements should be enforceable under the Dispute Settlement Understanding. Other parts of the agreement, such as rules on source-code disclosure or data localisation, should remain open for countries to join later if they choose.47
Second, the WTO must resolve the Annex 4 problem. There is now a plurilateral text that ninety-one participants negotiated and seventy-two members co-sponsor, but it cannot become part of WTO law. It therefore lacks both the legitimacy of a multilateral decision and the flexibility of a regional deal. Instead of helping, this text merely adds another strand to the web of overlapping trade rules it was meant to simplify.48
Third, any global baseline must allow for exceptions grounded in existing WTO rules, specifically GATS Article XIV as interpreted in cases such as China – Publications and US – Gambling. These carve-outs would let developing and least-developed countries protect their digital industries without being forced to accept high-level standards set by richer nations. Without such exceptions, smaller countries will have to choose between joining a trade regime on terms that favour powerful economies and staying entirely outside the rule-making process.49
Until WTO members recognise this, negotiations will continue in the dark. Every new regional deal adds another thread to a tangled knot, and only a shared multilateral baseline can fully unwind it.
That the United States withdrew its support for the core data-flow rules of the Joint Statement Initiative in October 2023, while those rules remain in the USMCA, suggests a troubling pattern: even the largest digital powers are unwilling to bind themselves to such rules through a multilateral body, preferring to enforce them against trading partners with less bargaining power.
*****
Footnotes
1. Agreement Between the United States of America, the United Mexican States, and Canada ch. 19, arts. 19.12, 19.16, Nov. 30, 2018 (entered into force July 1, 2020), https://ustr.gov/sites/default/files/files/agreements/FTA/USMCA/Text/19-Digital-Trade.pdf [hereinafter USMCA].
2. Digital Economy Partnership Agreement, modules 1–16, June 11, 2020 (entered into force for New Zealand and Singapore Jan. 7, 2021), https://www.treaties.mfat.govt.nz/search/details/t/3945/c_1 [hereinafter DEPA].
3. Comprehensive Economic Partnership Agreement Between the Government of the Republic of India and the Government of the United Arab Emirates ch. 9, art. 9.3(4), Feb. 18, 2022 (entered into force May 1, 2022), https://www.moet.gov.ae/documents/20121/1347101/Final+Agreement_UAE+India+CEPA.pdf [hereinafter India-UAE CEPA].
4. World Trade Organization, Declaration on Global Electronic Commerce, WTO Doc. WT/MIN(98)/DEC/2 (May 25, 1998).
5. World Trade Organization, Joint Statement on Electronic Commerce, WTO Doc. WT/MIN(17)/60 (Dec. 13, 2017); see also World Trade Organization, Joint Statement Initiative on E-commerce, https://www.wto.org/english/tratop_e/ecom_e/joint_statement_e.htm (last visited Oct. 24, 2026) [hereinafter WTO JSI Page].
6. Communication from the Co-sponsors, Agreement on Electronic Commerce, WTO Doc. WT/GC/W/955 (Dec. 2024); WTO JSI Page, supra note 5.
7. World Trade Organization, Members Review Possible Deliverables for MC14 at Final 2025 General Council Meeting (Dec. 16, 2025), https://www.wto.org/english/news_e/news25_e/gc_16dec25_255_e.htm; Peter Ungphakorn, E-commerce Initiative Is the Latest WTO Rule-Making Effort to Be Blocked, Trade β Blog (Feb. 19, 2025), https://tradebetablog.wordpress.com/2025/02/19/e-commerce-initiative-latest-blocked/; WTO JSI Page, supra note 5 (recording the Declaration on interim arrangements, WTO Doc. WT/MIN(26)/42 (Mar. 28, 2026)).
8. General Agreement on Trade in Services, Apr. 15, 1994, Marrakesh Agreement Establishing the World Trade Organization, Annex 1B, 1869 U.N.T.S. 183 [hereinafter GATS].
9. GATS, supra note 8, art. XIV(c)(ii).
10. Appellate Body Report, United States – Measures Affecting the Cross-Border Supply of Gambling and Betting Services, WTO Doc. WT/DS285/AB/R (adopted Apr. 20, 2005) [hereinafter US – Gambling]; Appellate Body Report, China – Measures Affecting Trading Rights and Distribution Services for Certain Publications and Audiovisual Entertainment Products, WTO Doc. WT/DS363/AB/R (adopted Jan. 19, 2010) [hereinafter China – Publications and Audiovisual Products].
11. China – Publications and Audiovisual Products, supra note 10, ¶¶ 242–244, 335–337; Brendan McGivern, Chinese Import Restrictions on Publications and Entertainment Products Found to Be WTO Inconsistent, 13 ASIL Insights no. 19 (Oct. 27, 2009), https://www.asil.org/insights/volume/13/issue/19/chinese-import-restrictions-publications-and-entertainment-products.
12. WTO JSI Page, supra note 5 (quoting World Trade Organization, Joint Statement on Electronic Commerce, WTO Doc. WT/L/1056 (Jan. 25, 2019)).
13. World Trade Organization, Information on the Agreement on Electronic Commerce ¶¶ 1.3, 1.15 (n.d.), https://www.wto.org/english/tratop_e/ecom_e/information_on_agreement_ecom.pdf (last visited Oct. 24, 2026) [hereinafter Information on the Agreement].
14. Henry Gao, The Joint Statement on E-commerce: Is This Glass Half Empty or Half Full?, Ctr. for Int’l Governance Innovation (Sept. 30, 2024), https://www.cigionline.org/articles/the-joint-statement-on-e-commerce-is-this-glass-half-empty-or-half-full/.
15. World Trade Organization, Declaration on Global Electronic Commerce, supra note 4; World Trade Organization, Work Programme on Electronic Commerce, https://www.wto.org/english/tratop_e/ecom_e/ecom_work_programme_e.htm (last visited Oct. 24, 2026) (recording that the Thirteenth Ministerial Conference, in March 2024, extended the moratorium until the Fourteenth Ministerial Conference or 31 March 2026, whichever was earlier, and that the moratorium lapsed on 30 March 2026).
16. USMCA, supra note 1, arts. 19.12, 32.1(2).
17. Michael Geist, How Canada Surrendered Policy Flexibility for Data Localization Rules in the USMCA, Michael Geist (Oct. 10, 2018), https://www.michaelgeist.ca/2018/10/how-canada-surrendered-policy-flexibility-for-data-localization-rules-in-the-usmca.
18. Zhihan Tian, Cracking the Digital Spaghetti Bowl: A Recipe for Streamlined Global Trade, Mich. J. Int’l L. Online (Nov. 2023), https://mjilonline.org/?p=3637.
19. Zakaria Sorgho, RTAs’ Proliferation and Trade-Diversion Effects: Evidence of the “Spaghetti Bowl” Phenomenon (Munich Personal RePEc Archive, Paper No. 60503, 2014), https://mpra.ub.uni-muenchen.de/60503.
20. Stephanie Honey, The Long Road to a Seamless Global Digital Economy, Hinrich Foundation (May 30, 2023), https://hinrichfoundation.com/research/article/digital/the-long-road-to-a-seamless-global-digital-economy.
21. Simon J. Evenett, Johannes Fritz & Tommaso Giardini, Deterring Digital Trade Without Discrimination, 117 AJIL Unbound 104, 104 (2023), https://doi.org/10.1017/aju.2023.15, also available at https://globaltradealert.org/reports/34.
22. Irina Korgun, South Korea’s Place in the Evolving Global Regulation on Digital Trade Architecture, 19 Int’l Orgs. Rsch. J. 70 (2024), https://doi.org/10.17323/1996-7845-2024-03-04.
23. Tian, supra note 18.
24. India-UAE CEPA, supra note 3, arts. 9.10, 9.11; DEPA, supra note 2, art. 4.4, Annex I; USMCA, supra note 1, art. 19.12; Economic Laws Practice, India – UAE Comprehensive Economic Partnership Agreements: Key Takeaways (Mar. 30, 2022), https://elplaw.in/india-uae-comprehensive-economic-partnership-agreements-key-takeaways/.
25. USMCA, supra note 1, art. 31.2; DEPA, supra note 2, module 14, Annex 14-A; Deborah Elms, Unpacking the Digital Economy Partnership Agreement (DEPA), Hinrich Foundation (Jan. 28, 2020), https://hinrichfoundation.com/research/talking-trade-blog/unpacking-the-digital-economy-partnership-agreement-depa.
26. India-UAE CEPA, supra note 3, art. 9.3(4); Economic Laws Practice, Trade Update (Apr. 18, 2022), https://elplaw.in/leadership/trade-update-april-18-2022/.
27. Digital Watch Observatory, The WTO Joint Initiative on E-commerce, https://dig.watch/processes/wto-ecommerce (last visited Oct. 24, 2026); Information on the Agreement, supra note 13, ¶ 1.3.
28. UNCTAD, Digital Economy Report 2019: Value Creation and Capture: Implications for Developing Countries (2019), https://www.doi.org/10.18356/3f10427a-en.
29. UNCTAD, The Digital Economy: A New Frontier for Trade, Investment and Sustainable Growth, U.N. Doc. UNCTAD/PRESS/IN/2025/005 (Oct. 17, 2025), https://unctad.org/system/files/press-material/in25005_digital-economy_en.pdf.
30. Diego Aulestia Valencia, Chair, Group of 77 and China, Statement at the Sixty-Eighth Executive Session of the UNCTAD Trade and Development Board, Item 5: Capturing Value in the Digital Economy ¶ 3 (Sept. 30, 2019), https://unctad.org/meetings/en/Presentation/tdbex68_item5_Ecuador_G77.en.pdf.
31. DEPA, supra note 2, modules 2, 4, 7, 8, 10; Press Release, Ministry of Trade & Industry (Singapore), Singapore Leads the Way in New Digital Economy Partnership Agreement with Chile and New Zealand (May 17, 2019), https://www.mti.gov.sg/Newsroom/Press-Releases/2019/05/Singapore-leads-the-way-in-new-Digital-Economy-Partnership-Agreement-with-Chile-and-New-Zealand.
32. Elms, supra note 25; Joint Press Release, Ministry of Foreign Affairs & Trade (N.Z.), Joint Press Release on the Accession of the Republic of Korea to the Digital Economy Partnership Agreement (June 8, 2023), https://www.mfat.govt.nz/en/media-and-resources/joint-press-release-on-the-accession-of-the-republic-of-korea-to-the-digital-economy-partnership-agreement; Joint Press Release, Ministry of Foreign Affairs & Trade (N.Z.), Joint Press Release on the Successful Accession of Costa Rica to the Digital Economy Partnership Agreement (Sept. 17, 2026), https://www.mfat.govt.nz/en/media-and-resources/joint-press-release-on-the-successful-accession-of-costa-rica-to-the-digital-economy-partnership-agreement.
33. Press Release, Ministry of Trade & Industry (Singapore), Digital Economy Partnership Agreement Joint Committee Commences Accession Working Group for China (Aug. 18, 2022), https://www.mti.gov.sg/newsroom/digital-economy-partnership-agreement-joint-committee-commences-accession-working-group-for-china; Joint Press Release on the Successful Accession of Costa Rica to the Digital Economy Partnership Agreement, supra note 32.
34. India-UAE CEPA, supra note 3, arts. 9.4, 9.8, 9.11, 9.16, 9.17; Economic Laws Practice, Key Takeaways, supra note 24; Digital Policy Alert, Signed: India-UAE Comprehensive Economic Partnership Agreement (CEPA) Containing Digital Trade Chapter, https://digitalpolicyalert.org/event/3579-signed-india-uae-comprehensive-economic-partnership-agreement-cepa-containing-digital-trade-chapter (last visited Oct. 24, 2026).
35. India-UAE CEPA, supra note 3, art. 9.15; World Trade Organization, Work Programme on Electronic Commerce, supra note 15.
36. Economic Laws Practice, Trade Update, supra note 26.
37. The Digital Personal Data Protection Act, No. 22 of 2023, India Code (2023), § 16; Ministry of Electronics & Information Technology, Notification G.S.R. 843(E) (Nov. 13, 2025), Gazette of India, Extraordinary, pt. II sec. 3(i); Jidesh Kumar, Data Localization vs. Cross-Border Flexibility: India’s Approach under the DPDP Act, 2023, KS&K (last updated Oct. 14, 2026), https://ksandk.com/data-protection-and-data-privacy/indias-dpdp-act-balancing-data-localization-flow.
38. Ahlawat & Associates, Opportunities Under the India UAE CEPA: Digital Trade, Duty-Free Access and Sector-Specific Benefits (Mar. 17, 2026), https://www.ahlawatassociates.com/blog/opportunities-under-the-india-uae-cepa-digital-trade-duty-free-access-and-sector-specific-benefits.
39. Information on the Agreement, supra note 13, ¶¶ 1.7–1.9.
40. World Trade Organization, Members Review Possible Deliverables for MC14 at Final 2025 General Council Meeting, supra note 7; Marrakesh Agreement Establishing the World Trade Organization art. X:9, annex 4, Apr. 15, 1994, 1867 U.N.T.S. 154.
41. Press Release, European Commission, EU-Singapore Digital Trade Agreement Enters into Force (Feb. 2, 2026), https://policy.trade.ec.europa.eu/news/eu-singapore-digital-trade-agreement-enters-force-2026-02-02_en; Honey, supra note 20; Digital Economy Partnership Agreement Joint Committee Commences Accession Working Group for China, supra note 33.
42. Information on the Agreement, supra note 13, ¶¶ 1.8, 1.11.
43. Kumar, supra note 37.
44. China – Publications and Audiovisual Products, supra note 10, ¶¶ 242–244, 335–337; US – Gambling, supra note 10.
45. USMCA, supra note 1, arts. 19.11, 19.12, 19.16; Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the Protection of Natural Persons with Regard to the Processing of Personal Data and on the Free Movement of Such Data, and Repealing Directive 95/46/EC (General Data Protection Regulation), 2016 O.J. (L 119) 1; Korgun, supra note 22.
46. Gao, supra note 14.
47. Information on the Agreement, supra note 13, ¶ 4.27; Elms, supra note 25.
48. World Trade Organization, Members Review Possible Deliverables for MC14 at Final 2025 General Council Meeting, supra note 7.
49. GATS, supra note 8, art. XIV; China – Publications and Audiovisual Products, supra note 10; UNCTAD, Digital Economy Report 2019, supra note 28.