Since 1998, when the World Trade Organization (WTO) adopted the Work Programme on Electronic Commerce (WPEC), the global landscape of the digital economy has undergone significant and long-term changes.
For one, over the last thirty years, the definition of ‘e-commerce’ has moved beyond straightforward digital advertising and now includes everything from social media platforms and online search engines and marketplaces to cloud computing services, seeing their highest demand at this moment. This expansion in the WPEC’s scope has turned some of its ‘free’ trade measures, like its customs duties moratorium on electronic transmissions, among others, into a flashpoint for trade negotiations. The recent Ministerial Conference (MC-14) and the General Assembly proceedings also indicate a growing challenge to the WTO’s legitimacy within this domain.
More importantly, the burgeoning “AI race” of our time no longer resembles the asset-light digital and platform economies of the previous decade.
More importantly, the burgeoning “AI race” of our time no longer resembles the asset-light digital and platform economies of the previous decade. The former’s demand for mature manufacturing industries has meant that the Global North’s historical control over financial capital and intellectual property (in software and chip design markets, for example) is no longer sufficient to sustain the massive investments being poured into the sector. At the same time, the emergence of competitive digital ecosystems in some Global South countries such as China, Brazil, India, and Kenya has also complicated, but not erased, the ability of American Big Tech in maintaining their market dominance across the supply chain.
However, an ineffective and partisan trade regime has meant that the development of these co-dependent relationships has failed to materialize into egalitarian ones. On the contrary, the threat posed to the US by China’s industrial rise has nudged the former towards a policy of state capitalism that treats AI as a critical sector of national security and economic importance. Washington’s repeated sanctions against China, its Section 301 enforcement against Brazil, and its use of reciprocal tariffs to bring India to the negotiating table are clear examples of this tendency. The US-led Pax-Silica Agreement and China-led World Artificial Intelligence Cooperation Organization (WAICO) further reiterate an emerging fragmentation of the global political order.
But is this landscape of a polarized information ecosystem the only option we have? The political economy of knowledge, which is at the core of the ongoing AI spring, is just another site for the interaction of labor and capital. So can Global South economies, particularly those with more mature digital economies, leverage their forces of production and distribution to challenge this wave of fragmentation, gatekeeping, and weaponization of the information commons? Is there a potential road not yet taken?
Reading the present: The ‘core’ is organized, the ‘peripheries’ are not
As mentioned above, the US’s role in disrupting an open and competitive internet ecosystem is undoubtedly significant. Even before the “AI race” truly took off, the Office of the United States Trade Representative (USTR) had been regularly objecting to developing countries’ efforts towards building local industries and markets. The National Trade Estimate Reports on Foreign Trade Barriers released by the USTR have constantly cited policies like digital services taxes, ex-ante competition regulation, national standardization requirements, and locally established digital public goods as unfair trade barriers that hinder the ability of American Big Tech enterprises.
Even before the “AI race” truly took off, the Office of the United States Trade Representative (USTR) had been regularly objecting to developing countries’ efforts towards building local industries and markets.
Although many of these reports challenged the actions of other Global North countries, including some of the US’s allies, it is important to note here that the capitalist ‘core’ continues to be on the same page on many principles of digital trade liberalization. The European Union (EU) and the United Kingdom (UK) have both, for example, recently inked trade agreements with India that continue to favor unrestricted cross-border data flows, oppose source code disclosures, and limit the countries’ ability to provide favorable treatment to domestic suppliers in public procurement processes. In many ways, this is contradictory to the rhetoric of territorial and national sovereignty that the two states demand for their local digital economies.
More important, though, is the relatively unorganized nature of digital and AI strategies being pursued by Global South peripheries. At the MC-14, for instance, Brazil was the only prominent internet economy from the Global South that resisted the US demand for a permanent customs duties moratorium on electronic transmissions; even its more prominent digital allies from the BRICS initiative failed to do so. India acceded to a four-year pause, whereas China remained a signatory to the Electronic Commerce Agreement (ECA)’s interim arrangements — alongside Australia, Canada, the EU, Japan, the UK, and other ‘core’ nations. Article 11 of the interim arrangements prevents the imposition of customs duties until a review after five years.
Without adequate and consistent safeguards at the domestic level, regulatory arbitrage is already becoming a competitive advantage for many Global South countries, which risks making many of these industrial activities extractive to labor and nature in the long run.
This policy misalignment within leading nations of the Global South, arising as a result of conflicting material realities and desires, is also highlighted within their AI strategies. Limited control over intellectual property and financial capital, and industrial capacity miles behind China, has restrained their economic options to data centers, second-order manufacturing, ‘sovereign’ models, language datasets, and AI applications. Without adequate and consistent safeguards at the domestic level, regulatory arbitrage is already becoming a competitive advantage for many Global South countries, which risks making many of these industrial activities extractive to labor and nature in the long run.
Nurturing futures — An open, neutral, and cooperative knowledge commons
From the open web to the ‘walled gardens’, the history of the digital economy provides us with two key learnings about building a global ecosystem of information on complex software and hardware supply chains.
The enclosure of data further narrows the scope and quality of public interest innovation.
First, data — one of this ecosystem’s primary inputs — is a non-rivalrous and non-excludable good, which means networked markets centered around collecting and processing more of it are highly vulnerable to private capture and state surveillance. The enclosure of data further narrows the scope and quality of public interest innovation. We need guardrails to ensure the value of data is effectively ‘commoned,’ and this needs brave out-of-the-box measures, such as mandatory training data frameworks committed to harm prevention and baseline norms for the economic governance of data to enable equitable benefit sharing.
Secondly, and more importantly, international and multilateral alignment over technical standards alone is insufficient to counter these outcomes, thereby demanding that nation states begin identifying and adopting common legal principles. This is especially true for the AI economy, where transnational LLCs use regulatory arbitrage, open-washing, decentralized infrastructures, and appeals to national sovereignty to side-step domestic rules that incentivize cooperative innovation, democratic accountability, and equal distribution of the intelligence dividend.
It is not enough to argue about what constitutes openness and interoperability in the AI stack, but to recognize that the question here is of ensuring the “publicness” of foundational data and AI infrastructures, and creating techno-institutional policy measures that build an empowering dynamic between digital public goods and the innovation commons.
It is not enough to argue about what constitutes openness and interoperability in the AI stack, but to recognize that the question here is of ensuring the “publicness” of foundational data and AI infrastructures
Indeed, international organizations like the International Telecommunication Union (ITU), the Internet Engineering Task Force (IETF), and other standard-setting bodies provide valuable lessons in building technical specifications collaboratively, be it on model openness, interoperable compute, or data provenance. Similarly, the rules and frameworks put forth by the UN, its affiliates (WTO, WHO, WIPO, UNICEF, and World Bank), and other multilateral bodies reflect crucial precedent in setting cross-border principles for AI governance. But a commons-based realignment would be impossible without dealing with more material concerns of collective ownership, pluralistic governance, and a just distribution of resources.
In addition to market-based routes for technology transfer (like IP exemptions for GPU manufacturing), we need regional and South-centric cooperation partnerships for the effective creation of technological capabilities in the Global South. For instance, as this joint statement by the Global Digital Justice Forum and the Global South Alliance notes, developing public compute infrastructures with provisioned access controls at national levels could be a key objective for international financing and regional development collaborations. Similarly, sustained inter-governmental coordination and knowledge sharing between competition, labor, environment, and tax regulators would improve mandate enforcement across countries.
And while trade agreements have, traditionally, not been satisfactory avenues for such coordination, a South-centric repositioning could present a better alternative to the conventional trade regime.
And while trade agreements have, traditionally, not been satisfactory avenues for such coordination, a South-centric repositioning could present a better alternative to the conventional trade regime. The African Continental Free Trade Area’s (AfCFTA) Digital Trade Protocol, for example, allows member states to enforce customs duties on digital products based on specific rules of origin.
This article emerged through the author’s interaction with digital trade researchers and civil society actors at a conference organized by the Global Fund for a New Economy. The author would like to thank the Fund for the invitation and the participants for their constructive engagement with the arguments presented above.
