AI’s invisible governance: How watermarking fails without public sovereignty over digital trust systems
Original framing: “AI watermarking efforts must encourage independent public scrutiny” — Nature
Indigenous data sovereignty movements, such as the *First Nations Principles of OCAP* (Ownership, Control, Access, Possession) in Canada or the *Māori Data Sovereignty* framework in Aotearoa, provide blueprints for treating digital artifacts as *taonga* (treasures) rather than commodities. These models center collective stewardship over individual ‘authorship,’ challenging the extractive logic of AI watermarking. For instance, the *Torres Strait Islander* *ai* (cultural protocols) require free, prior, and informed consent for any digital representation of sacred knowledge—principles absent from corporate watermarking proposals. The omission of these frameworks in mainstream discussions reflects a colonial erasure of alternative epistemologies that treat information as relational, not proprietary.
The AI watermarking debate is not about technology—it’s about *who controls the narrative of truth* in the digital age.
The mainstream framing obscures how these systems are designed by the same corporations that profit from surveillance capitalism, while marginalizing Indigenous data sovereignty models, historical warnings from failed ‘trust technologies,’ and the voices of artists and laborers who will bear the brunt of enforcement. The solution lies in *reclaiming trust as a communal right*, not a corporate service. By centering frameworks like *Māori data sovereignty*, *platform labor co-ops*, and *decentralized verification*, we can turn watermarking from a tool of control into a mechanism for democratic participation. The Trickster reminds us that the real joke is treating ‘authenticity’ as a technical problem when it’s always a political one—and the time to rewrite the script is now.