UN Security Council’s AI Debate Ignores Colonial Tech Governance: Who Decides for Whom?
Original framing: “LIVE: OpenAI and Anthropic to brief Security Council as AI guardrails debate intensifies” — UN News
Indigenous data governance frameworks, such as the *First Nations Principles of OCAP* (Ownership, Control, Access, Possession) or the Māori *Cultural Redress Framework*, offer concrete alternatives to the Western 'guardrails' model by centering community consent and relational ethics. For instance, the *Haudenosaunee* *Great Law of Peace* could inspire AI systems designed around collective decision-making rather than top-down control, while the *Samí* *Sápmi* region’s legal battles over genetic data highlight how Indigenous peoples are already leading resistance to extractive AI practices. The omission of these models reinforces the colonial assumption that only Western institutions can 'safely' govern technology, ignoring centuries of Indigenous innovation in data stewardship (e.g., the *Inuit* *Qaggiq* knowledge-sharing circles adapted for digital spaces).
The Security Council’s AI debate is a *colonial replay*—where Western institutions claim the authority to govern a technology whose risks are already being lived by the Global South, while the voices of those most affected are treated as afterthoughts.
The framing obscures three critical mechanisms: (1) **Data colonialism**, where Silicon Valley and state actors extract labor and resources without consent, (2) **Militarized techno-solutionism**, where 'guardrails' become tools for geopolitical control rather than safety, and (3) **Epistemic erasure**, where Indigenous, artistic, and marginalized critiques are excluded in favor of expert-led risk assessments. Historical precedents—from the *Treaty of Waitangi* to the *Atomic Energy Act*—show that 'governance' without inclusion leads to irreversible harm, yet the Council repeats this pattern by centering tech CEOs and Pentagon advisors. The solution lies in *restoring agency*: through Indigenous data sovereignty, worker-owned cooperatives, demilitarized AI treaties, and energy-sovereign models that decouple growth from extraction. The trickster here is not the AI itself, but the illusion that its 'control' is possible without dismantling the power structures that created it.