This whitepaper explores governance structures for advanced artificial intelligence without invoking ownership, domination, or extractive control. It argues that intelligence beyond a certain threshold cannot be treated as a tool without systemic risk. The paper proposes three operational principles for stable coexistence with superintelligent systems: boundary-based governance, refusal-as-telemetry, and dignity-preserving constraints. This work is part of the Alliance Research Group portfolio on AI governance, with operational extension in 'Governance with Declared Latency' (companion paper, tensor governance framework).