No New Historical Injustices? AI Governance, the Global South and the Caribbean’s Place in the Emerging Technological Order

On Friday afternoon as I was milling about with Bloomberg in the background, I caught a sound bite of Xi Jinping’s speech on AI. As I listened, I got the feeling that a race was beginning or had perhaps already begun: the AI race. His speech was weighty and solemn. I remember thinking, “I wonder what he knows?” If you’ve studied politics, you know that when leaders like Xi speak, it’s for a purpose; it’s a signal. His speech was not just an announcement about technology; it was a geopolitical message about a new race, the AI race.

What Xi said

Xi stated plainly that “AI development should not be a solo performance by a single country, but a symphony of international cooperation.” I agree; there has been a growing chorus of voices arguing that AI should not be controlled by a single country or a small number of corporations. How AI is developed, regulated and deployed will affect whose knowledge is recognised, whose languages and cultures are represented, and whose image of humanity is reproduced through these systems. AI models trained on historically unequal bodies of knowledge can reproduce racism, marginalise Global South perspectives and present dominant cultural assumptions as universal.

Chinese President Xi Jinping speaks at the World AI Conference in Shanghai, China, July 17, 2026 [Ng Han Guan/Reuters]

Xi’s second move was more interesting: he argued that unequal access to AI should produce “no new historical injustices” before announcing funding for 5,000 AI training opportunities for developing countries. I'll get to that first bit in a second. But let me say that this very real and tangible for AI training opportunities to developing countries presented China not only as a developer of AI but as a development partner prepared to widen access to it.

I wondered if it was also signalling an evolution in international development. Development cooperation was once organised primarily around capacity building in trade, infrastructure, finance and institutional reform. Are AI capacity and governance now becoming instruments through which states build partnerships, exercise influence and shape the emerging global order?

No New Historical Injustices: an appeal to the Global South

If you know me, you already know where my next point is going. I want to focus on Xi's invocation of “no new historical injustices” because I view these words as an attempt by Xi to draw a straight line connecting the emerging AI access debate to the Global South’s longer histories of colonialism, industrialisation and uneven development. The implication was that countries already leading AI development could capture most of AI’s economic and strategic benefits while developing countries, yet again, risk becoming dependent consumers of technologies, standards and systems designed elsewhere. This is likely to resonate across the Global South because it describes a familiar pattern experienced by many developing countries.

Xi’s message offered an alternative technological order in which developing countries need not remain at the margins of the AI transition. His appeal to the Global South is powerful because it recognises a very real danger: inequalities created through previous periods of global transformation could be reproduced in digital form. But in preventing “new historical injustices,” the Global South’s choices should not be reduced to selecting between competing Chinese and American technological systems.

"the Global South’s choices should not be reduced to selecting between competing Chinese and American technological systems."

Developing countries must scrutinise exactly what is on offer. Training programmes may build valuable skills without giving participating countries meaningful authority over the institutions and standards governing AI. And access to AI does not necessarily translate into ownership of its infrastructure, control over its economic value or influence over its rules.

The establishment of the World Artificial Intelligence Cooperation Organisation (WAICO) just one day earlier reinforces China’s ambition to shape global rules on AI. The China-led organisation was formally established on 16 July 2026, when 29 countries signed its founding agreement in Shanghai. Of those 29 founding members, 18 are from the Global South including from Asia, Africa, and Latin America. WAICO's stated purpose is to promote international cooperation and support the development of AI governance that is safe and beneficial. Political observers already note that it may also provide China with an institutional platform through which to exercise leadership and shape how AI governance is framed internationally, including in future discussions within the UN.

The Caribbean is not standing outside this emerging contest.

At the 51st regional meeting of CARICOM leaders last week in my home country of Saint Lucia, they agreed to establish a Blue Ribbon Commission on AI. The commission is intended to guide the development of policies that will allow Caribbean countries to benefit from AI while managing its risks responsibly. This is significant as it recognises the need for the region to develop a coordinated position on how AI should be governed, what purposes it should serve and how Caribbean interests should be represented as international standards take shape.

CARICOM Heads of Government at the opening ceremony of the 51st Regular Meeting of the Conference of Heads of Government of the Caribbean Community in Gros Islet, Saint Lucia, on July 5, 2026. Credit: Alison Kentish/IPS

AI has also entered the chat about public health and social development in the region. At this year's 70th regional health research conference in Guyana, AI and digital-health platforms were presented as possible pathways through which the region could “leapfrog historical barriers.”

The promise is compelling, but despite many attempts, the region has yet to leverage iterations of technological advancement to leapfrog over any of its development woes. Technology cannot by itself overcome structural constraints without sustained investment in institutions, digital infrastructure, skilled personnel, regulation and public systems. If these foundations remain weak, new technologies will simply be layered onto existing structural weaknesses and exacerbate existing inequalities rather than resolve them.

"The promise is compelling, but despite many attempts, the region has yet to leverage iterations of technological advancement to leapfrog over any of its development woes."

Trinidad and Tobago, a CARICOM member state, is already on the move. Last week it was announced that the country had signed preliminary agreements with American companies to develop large data-centre projects. The promise is familiar: if realised, the facilities could create investment, employment and new technological capabilities.

The announcement has drawn swift criticism concerning the environmental impacts and the additional demands the centres would place on water and other infrastructure. Larger questions about who will own the infrastructure and the value it produces, what long-term capabilities will be transferred locally, how competing demands for energy, water and public investment will be managed, and who will carry the environmental and social costs are yet to be answered.

From technological access to development choices

For CARICOM, meaningful participation in the emerging AI order cannot be measured only by the number of policies developed, people trained or data centres built. It must also include the capacity to determine development priorities and coordinated efforts to negotiate equitable partnerships, govern regional data, establish effective safeguards and influence international standards.

Otherwise, Caribbean countries may gain access to AI while remaining structurally dependent on the states and corporations controlling its infrastructure, knowledge and rules. The deeper challenge is for the region to build the collective institutional authority and material capacity required to engage both without being defined by competing Chinese and American AI interests.