<p>This article addresses a fundamental philosophical problem in environmental governance: Can non-conscious artificial intelligence systems provide legitimate representation for non-conscious ecological systems within democratic frameworks? I argue that the epistemic asymmetry between human cognition and ecosystem functioning creates what I term "double non-conscious representation"—AI systems representing ecological entities that, like the AI systems themselves, lack consciousness in traditional senses. This arrangement requires reconceptualizing both representation theory and democratic legitimacy beyond anthropocentric assumptions. Through analysis of ecological commons governance as a paradigmatic case, I develop a framework of "technological constitutionalism" that identifies conditions under which AI guardianship can enhance rather than undermine democratic environmental governance. The core philosophical contribution lies in demonstrating how epistemic privilege regarding ecological systems can ground bounded autonomous authority for AI representatives despite the absence of conscious agency in both representative and represented, provided appropriate institutional constraints preserve democratic accountability while enabling genuine ecological decision-making.</p>