Artificial Intelligence is rapidly becoming part of the infrastructure and economy of countries around the world. But as countries adopt AI, questions are growing about who owns and controls the technology, the data it relies on, and the decisions it increasingly shapes.
For Pacific countries, these questions are particularly important. Most will not aim to build frontier models or large-scale compute infrastructure in the near-term. But AI will still shape core development outcomes across the region, from disaster response and climate adaptation to health and education delivery. Many of the themes sure to be included in the Pacific Digital and Artificial Intelligence Technology Principles endorsed by Pacific leaders in Palau last week are centred around an important issue for the region: AI sovereignty.
AI sovereignty is about more than where servers are located. Sovereignty extends from compute infrastructure and data centres down to models and applications. It also means having the skills to understand and govern AI; to be able to protect local data and preserve cultural knowledge appropriately. Without that control, AI risks becoming another channel through which dependency deepens rather than one through which Pacific countries drive their own development.
Development partners and organisations are turning their focus towards supporting sovereign AI in the Pacific. But we must listen and understand if we’re really responding to the Pacific’s needs.
This week for the Intel we asked three experts: AI sovereignty – what matters for the Pacific?
AI sovereignty is more than just an abstract idea. It requires expensive physical components, infrastructure, and local capacity to run and manage such centers. In our region of mostly Small Island Developing States (SIDS), we lack both infrastructure and capacity. Our small island nations are still building basic ICT infrastructures and are far from moving our gaze to AI infrastructure. We need to be realistic about what we have to offer and what our limitations are. For now, we only have our data, our unique Pacific values and traditional knowledge, and our combined Blue Pacific Continent. This doesn’t mean we shouldn’t be thinking about AI sovereignty, rather, we should be at the table as stakeholders and approach it together and collectively as a region.
Regionalism is our best strategy forward. With AI becoming the next frontier for competition in our region, our leaders need to approach any AI solutions and investments cautiously to avoid contracts that conflict with our long-term efforts in combating climate change. After all, AI is the most energy-hungry technology we’ve ever seen.
Not only does AI pose existential risks, like draining our natural resources at an ungodly pace, it also poses real immediate risks. In Palau, right after the new fiber optic cable went live in 2017, we saw the rise of what is now referred to as technology-facilitated-gender based violence (TFGBV). Intimate photos and videos of community members were spread online, and communities were unprepared for the public proliferation and normalisation of sex acts considered taboo. Now with AI, TFGVB is in overdrive, with nefarious actors using generative AI to undress women and children. AI sovereignty discussions must address both the everyday and existential risks we face.
Joleen “Jo” Ngoriakl is an indigenous Palauan policy and international development practitioner with over a decade of experience designing and implementing governance and online safety initiatives. She is the founder of Palau Digital Citizenship Initiative and a UNIDIR Women in AI 2026 Fellow. At the Lab, we love Jo’s ability to bring cultural knowledge and customs to the front of emerging technology issues.
While AI sovereignty does not require the Pacific to build or own every AI technology from scratch, Pacific nations cannot afford to adopt external systems built on assumptions that overlook our languages, cultures, environments, and priorities. This risks allowing designs, stories, and environmental wisdom to be used in ways that do not reflect our identity, such as increasingly widespread AI-generated products like AI Polynesian art and tattoos. The Pacific must strengthen frameworks that uphold consent, cultural integrity, and equitable benefit-sharing.
Pacific nations are already shifting to pool political and technical capacity, and advance region wide technology principles through a Regional AI Governance Framework, joint AI toolkits, shared Digital Public Infrastructure (DPI), and a proposed Pacific Islands AI Technical Assistance Facility. This facility would provide a central expert network to support negotiations with global providers, serve legal expertise, workforce development, and cultural protection.
Donor support can help Pacific governments establish the policy and legislative foundations for technical capability and workforce development.
Australia can better support the Pacific by investing in long‑term capacity instead of short‑term, fly‑in fly‑out projects which disrupt momentum and place pressure on already stretched local teams. Remote technical assistance and maintaining data infrastructure locally can also build continuity and reduce dependency.
Although significant attention is given to SMEs and entrepreneurship, gaps also remain in supporting Pacific tech experts, particularly those outside government and dependent on project-based work. Without these specialists, countries lose out on hard‑won expertise. Engaging and resourcing local talent through ongoing projects is one of the most effective ways to retain skills, strengthen national capacity, and reduce brain drain.
Ultimately, AI sovereignty is a pathway to Pacific self‑determination, cultural integrity, and community power, and it starts with investing in Pacific expertise.
Cherie Lagakali is a Pacific digital inclusion and online safety specialist working at the intersection of technology, policy, and community resilience. As Founder of Kairos Kapacity NZ, she leads a Pacific‑centred storytelling and digital confidence platform uplifting women and underrepresented communities. At the Lab, we love Cherie’s ability to deliver practical, contextually grounded advice and high‑quality analysis.
There are two main elements when it comes to AI Sovereignty. The first is about data access and compute power, including the types of LLM models, the data used to train the models and the capacity of the infrastructure. Second is about the Sovereignty, the rule of law, and the extension of a national line of commitment to safeguard our way of living.
For the Pacific, AI sovereignty is about preservation of the Pacific inheritance and way of life. For many years we have lived in harmony with our ways and customs. We must focus on protection of our roots and customs as we build new Digital Islands using the Cyberspace that we are forced to accept to survive as nations. The questions we must answer are: who should we trust? How should we work together as a regional family to build that trust? Where do we build this trust together to ensure our islands are not exploited for a return on investment?
Australia could support a shared Pacific AI Digital Hub alongside Digital Embassy Facilities that would allow Island states to develop AI capabilities together while maintaining access to Pacific data and information. A regional hub would include experts from pacific nations and likeminded partners in aligning national and Pacific government needs. It would also reduce energy, staffing and technological costs compared to bilateral development and engagement. Through regional collaboration, we can build the safeguards to protect our values and customs when interacting with and building AI capabilities.
Suetena Faatuuala Loia is the Managing Director and Chief Technical Officer of Bytewatchers. He is an AI consultant, cyber security consultant and digital transformation enabler based in Samoa. The Lab loves Suetena’s sharing of lessons from his extensive practical experience at the forefront of digital transformation in the Pacific region.
