Data sovereignty is youth sovereignty: Rewriting the rules of AI and open data

Originally published in Manila Bulletin

AI promises to “know” everything—but who decides what can be known, and by whom? As a young leader, I am honored to be part of One Young World’s Indigenous Advisory Circle and to work alongside thousands of Indigenous youth who are advancing data sovereignty in our communities. For us, data is not raw material. It carries story, ceremony, and responsibility. When algorithms scrape the internet and call it “open,” they often ingest our culture without consent. The fastest way to make AI safer is also the most ethical: let Indigenous peoples govern our own data—starting with youth who will live with the consequences the longest.

Indigenous Data Sovereignty means our peoples decide what information about us—our lands, languages, kinship, and knowledge—gets collected, how it is stored, who can use it, and under what conditions. Consent is not a checkbox; it is an ongoing relationship with obligations on both sides. In practice, that looks like community-controlled rules for collection, access, use, sharing, and benefit across the full data lifecycle, with the authority to say yes, no, or not yet.

When our data is treated as ownerless, harm follows. Sacred songs are mislabeled as “folklore” and fed into models that remix them into novelty. Location and ecological records expose sacred sites and hunting grounds to trespass and exploitation. Health or justice datasets, shared without governance, become tools of surveillance. “Open” is mistaken for “free to take,” and our languages and teachings are used to train systems that return nothing to the people who created them. Even well-intentioned platforms erase us by standardizing our names and territories until identity is rounded off for the sake of accuracy.

Across One Young World’s global community, Indigenous youth are already building a better path. Community data trusts hold datasets under our own governance so access can be time-bound, purpose-limited, and revocable. Traditional Knowledge Labels travel with cultural materials in digital spaces, signaling protocols such as attribution, seasonal use, or restrictions on derivatives, so context and consent remain visible. Free, prior, and informed consent applies to datasets just as it does to land: before collection or training, proponents explain purpose, risks, benefits, and alternatives in the right language and on our timelines, and communities decide. In language technology, young developers co-create keyboards, speech tools, and archives with elders, keeping audio under community keys and ensuring any future model training returns benefits to the language community. Guardians and rangers use sensors and drones to protect rivers and forests while masking sensitive geolocations and encrypting stores so raw data cannot be scraped into external systems. These are not hypotheticals; they are practical safeguards that let innovation move faster because trust is built in.

Now the responsibility shifts to those who profit from data. If you build models, platforms, or “open” repositories, adopt binding governance agreements with recognized Indigenous entities, with clear purposes, retention limits, review rights, and remedies. Do not train on Indigenous content without documented consent, and honor “do-not-train” defaults and TK Labels technically, not just rhetorically, with post-training removal pathways that actually work. Share value through revenue, jobs, scholarships, and paid roles for community members, especially youth, in research, engineering, and governance. Fund community-controlled infrastructure—data trusts, local servers, connectivity, and Indigenous data labs—so we are not dependent on your cloud to access our own materials. Build cultural safety into release processes, with Indigenous youth involved in red-team exercises and the power to halt launches when risks outweigh benefits.

Governments and funders must keep pace. Make FPIC for data the law, procure only systems that demonstrate compliance with Indigenous governance where relevant, and back multi-year, community-led programmes that train youth in data science, governance, and cybersecurity alongside Elders’ knowledge.

I write this as someone who sees daily the ingenuity of our peers across continents. We are not asking for a pause on progress. We are offering a way to build technology worthy of the world it shapes. Data sovereignty is youth sovereignty. If you want AI that is safer, fairer, and more useful, put Indigenous youth at the table with the authority to govern our data and resource the infrastructures that make that authority real. The choice is simple: continue extraction 2.0, or move with us to reciprocity, consent, and shared benefit.

This opinion column is published under the Creative Commons Attribution 4.0 International License (CC BY 4.0). You are free to share, adapt, and redistribute this content, provided appropriate credit is given to the author and original source.