THE PATHWAYS OF ARTIFICIAL INTELLIGENCE AROUND THE WORLD AND THE RISKS FOR PACIFIC ISLAND STATES

PC: COLON FRELD

“The islands were, in fact, so to speak, experiments of nature, places which, by blessing or curse, were compelled by their geographical singularity to harbour unique forms of life. All singular species or genera which, in their isolated habitats, had embarked upon separate evolutionary trajectories.”

Neurologist Oliver Sacks’s warning (in “The Island of the Colourblind”, Adelphi) regarding the Micronesian atoll of Pingelap also serves as a cautionary tale for the relationship between Pacific communities and AI. In tourist brochures, these are lands with azure hues, variegated streaks of clouds, idyllic paradises; but people have long since stopped believing in postcard images, partly due to the growing coastal erosion caused by rising sea levels which, in cases such as Tuvalu, will lead to their disappearance. Nature’s intervention in the Pacific is set against the backdrop of a war with undefined contours: artificial intelligence points to an ‘evolutionary trajectory’ which, in this ‘geographical singularity’, highlights an unbalanced relationship.

On the one hand, small island nations scattered across the ocean, comprising of distant and sparsely populated atolls, which have long remained on the margins of the major changes of recent decades – widespread industrialisation, rapid urbanisation, mass consumerism, and rifts in intergenerational relations. On the other hand, sophisticated technology, offered in various forms by both private and public entities, which finds in the Pacific an almost ideal testing ground for experimentation and dominance. The slow pace of this part of the world, based on a purely community-oriented structure, stands in stark contrast to the rhythm dictated by the speed of technological evolution.

Let us consider three figures from the 2025 UNCTAD reports (https://unctad.org/publication/technology-and-innovation-report-2025): 40% of AI development is carried out by just 100 companies, all located in China and the United States, the two Pacific superpowers that can invest most easily in this region; 118 countries are completely excluded from AI governance – among these are the Pacific islands, which, in practice, must therefore accept what is offered to them and have little say in the process itself.

It should also be noted that there are regions which, due to their instability and the underdevelopment of their basic infrastructure, are unlikely to benefit from AI, thereby widening the development gap between them and others. In Asia, Timor-Leste, a small country and one of the poorest in the region, situated between two regional giants - Indonesia and Australia - represents an almost perfect case for attracting and importing smart technologies. On the other hand, Afghanistan is currently completely cut off from them and will fall further and further behind.

As for investment in research, it is estimated (though this is no easy calculation, as many programmes are mixed and many are classified because they relate to the military sector) at around €4 trillion, and this entails a reduction in spending in other areas of cooperation. Instead of education and basic infrastructure, clinics and scholarships, the future of foreign policy influence in the Pacific islands lies in the use of AI across numerous sectors, satellite networks, undersea cables, digital connections, access to databases, and social networks. The American lesson from USAID speaks volumes, and the Pacific is one of its first testing grounds. In this acceleration of time in the face of a fragmented and immense expanse, we shall describe these seas with words different from those used by Nobel laureate Derek Walcott for his islands:

“The amen of calm waters,

The amen of calm waters,

The amen of calm waters.”

These now turbulent waters are also European. There is a part of Oceania that belongs to France; there are strategic agreements with Japan, New Zealand, Australia and others, which are also being implemented to address Chinese expansionism and the new US approach in the region from an AI perspective. There is the ‘European Strategy for the Indo-Pacific’ and the resources of the Global Gateway, which provide a legal basis and funding mechanisms (among other things, the European Investment Bank opened an office in Fiji in 2023) for comprehensive cooperation that includes artificial intelligence, which in the Asia-Pacific region takes the form of humanitarian aid, smart cities, supply chain optimisation, pest and disease management, food quality control, inclusive economic growth and learning, energy efficiency, effective disaster response, sustainable agricultural practices, climate resilience and environmental protection, intergenerational connections and well-being.

Each of these areas requires specific professional expertise, with a growing knowledge gap, particularly in the Pacific, not only among the beneficiaries – the end-users and the island leadership – but also among those who decide on and coordinate such projects in various capacities.

Although it is sometimes difficult to admit, this is a sensitive issue that needs to be raised among scientists, engineers and legislators, as well as diplomats and civil servants, who must take action on matters where, to a large extent, many understand the broader picture but have only a limited grasp of the details of specific projects – knowledge that is the preserve of a select circle of experts. A risk that is even greater in the Pacific than elsewhere.

The risk is linked to the type of product, or more precisely to the supplier. AI is not a one-size-fits-all solution, and there are at least three ‘user manuals’ available.

China is largely developing a model based on state-subsidised and coordinated production, with public investment and objectives set by the central government to secure partnership agreements with third countries. Products for telehealth, tele-education, remote early-warning systems for potential natural disasters, and much more, become tools deliberately incorporated into a framework of foreign policy, bilateral cooperation and friendship. The orchestration of these partnerships follows a precise logic, a plan to establish a foothold in a specific country, with the flexibility needed to adapt to local needs and to maximise influence.

The Pacific region is being encouraged to adopt Chinese solutions, through agreements negotiated by national capitals and product packages designed for integrated, sector-by-sector development.

The United States is pursuing a more laissez-faire approach, relying primarily on businesses to develop and promote products for new markets. The 2023 US-Pacific Forum was a milestone in fostering relations between the business world and island nations. AI is developing hand-in-hand with essential digital connectivity and undersea cables. The public sector can intervene with financial support schemes, but the real driving forces appear to be industry leaders, who find the Pacific to be an open-air laboratory in which to test the effectiveness of certain new technologies.

The European Union has chosen a different path, typical of our part of the world: marked by rules, rights and progress targets. AI is a tool for pursuing greater environmental sustainability – such as the regional Team Europe initiative on the BlueGreen Alliance with the Pacific islands, which focuses on climate action, resilience and the responsible use of natural capital.

With the adoption in 2024 of the Artificial Intelligence Act, the first cross-sectoral legislation in this field, the EU has established that new technologies cannot be used for social scoring based on ethnicity, origin, sexual orientation or other personal data, nor can they be used by companies to measure the efficiency of their employees; there are also restrictions on their use by law enforcement agencies and strict limitations on the mandatory use of facial recognition, as well as safeguards for the confidentiality of personal data. A body of regulatory constraints that sometimes hinders the product’s competitiveness, in terms of the speed of project implementation and the achievement of real objectives by public authorities.

The European vision seeks to anchor AI in an approach centred on humans and their needs, and to curb potential abuses of AI, its potential to shape dystopian worlds. After all, for many civilians, AI is primarily a set of military technologies that leave no room for escape: the Russian aggression against Ukraine is claiming many victims due to atrociously ‘smart’ weapons, just as in the Gaza Strip there are numerous crimes against humanity caused by Israel’s ‘smart weapons’.

In this rapidly evolving landscape, following the European Union’s Artificial Intelligence Act, other countries and regional bodies – Brazil, India, ASEAN and the African Union – have begun to develop their own frameworks for the use of AI. It is never too early to start.

Efforts to coordinate the various approaches have so far been modest, but it is worth noting that in September 2024 the Artificial Intelligence Summit was held, attended by the United States, the United Kingdom, the European Union and many others, and that a binding treaty was signed; however, this is a treaty of limited scope as it is confined to monitoring the situation and does not provide for the implementation of concrete measures. 

Subsequently, in February 2025, chaired by French President Macron and Indian Prime Minister Modi, the Artificial Intelligence Action Summit was held in Paris, attended by more than 100 countries, including the European Union and many others, mainly from Asia and the Pacific. Fifty-eight of these countries signed a declaration on inclusive and sustainable artificial intelligence. Unfortunately, key players for the European Union, such as the United States of America and the United Kingdom, chose not to sign this declaration, confirming that even within the Western world, a shared vision is still lacking.

Furthermore, although the European Union has agreed to allocate around €200 billion to AI, Europe also suffers from significantly lower levels of investment compared to the United States and China, as well as being geographically further removed from the Pacific. This highlights the need to forge partnerships with those countries in the region that generally share the basic parameters for the development and use of AI – namely Japan, Australia, New Zealand and South Korea. In addition to research and the continuous, rapid introduction of new products, AI therefore thrives on increasingly frenetic institutional activity, characterised by divergent visions, limited agreements or large-scale partnerships, industrial strategies, and spheres of influence: power.

In this context, the Pacific represents a macro-region as vast as it is delicate. Its geographical and social characteristics constitute an ideal environment for the remote introduction of new technologies; yet, moreover, the Pacific nations will never possess sufficient political clout to negotiate fair AI agreements with major suppliers: Nauru has fewer than 15,000 inhabitants, whilst around 130,000 are scattered across the twenty or so inhabited islands in the three archipelagos of Kiribati; everywhere, the populations are small in number, the geography is vast and fragmented, and public infrastructure is meagre – and all of this far from prying eyes and public scrutiny.

These are the ideal settings for introducing (and experimenting with) new forms of AI.

Small island communities can thus benefit from telemedicine services that consider medicine delivery times, climate, availability and dietary habits, social, demographic and care data, and family structures, and provide dedicated remote medical services covering both general and specialist care. The same applies to the education of students through virtual teachers and teaching materials provided on the basis of algorithms that identify the community’s precise needs, or to support for local crops through seeds and special fertilisers designed to counteract the negative effects of the climate or the terrain, and also to connectivity with public administration services from which, until now, more or less remote island populations have been excluded. At a pace that may be much faster than anticipated, AI can transform the conservatism and geo-social characteristics of Pacific communities by offering services and unlocking possibilities hitherto unthinkable.

But who pays? And why, above all, in the Pacific?

AI is not a right, not even in principle, like access to water, literacy or internet connectivity. AI is a revolution created by visionary scientists, developed by companies that, by their very nature, seek profit, and utilised by states that can afford it, not only domestically but also as a new tool of foreign policy. AI is not, particularly in the Pacific, a gift, but an opportunity to forge partnerships that go hand in hand with control, access to data, and the evaluation of results obtained in a region with unique characteristics, with a view to further research and new products to be offered on a larger scale in broader markets. It is therefore not surprising that it is precisely the European approach – less cynical, more regulated and scrupulous – that is lagging behind elsewhere, such as in the Pacific, which more than anywhere else bears the hallmarks of the Wild West.

The gap that is opening with the forms of development cooperation and donor aid known to date (which are always self-serving) goes hand in hand with the dependency that AI creates in small island nations. A well or a clinic remains even after the donor has moved on, but an algorithm and a smart digital infrastructure are bound to create a need that will be difficult to do without once the transition is complete.

In the unknown scenarios that lie ahead, the possibility of moving away from AI and into a post-AI era remains uncertain and mysterious – particularly in small communities cradled by the vast and ever-warming ocean, due to geostrategic interests and climate change. In such a vulnerable and needy context, the revolution in possibilities and services offered is exceptional, but so is the risk: lacking bargaining power with industries, in an impossible balance of power with the major powers, and outside the realm of the technical knowledge essential to exercise any form of control, artificial intelligence offers greater emancipation whilst simultaneously taking away freedom. In the 21st century, this is called neo-colonialism, even neo-slavery.

Niccolò Rinaldi

Head of the Asia-Pacific Unit at the European Parliament and former MEP

---