AI互操作协议治理模式比较：DAO与公司主导标准的可持续性实证研究

arXiv:2606.05667v2 Announce Type: replace Abstract: As artificial intelligence (AI) agents enter production infrastructure, interoperability protocols shape its governance and sustainability. This paper revisits our comparative study of two AI-agent interoperability standards, Ethereum Request for Comments 8004 (ERC-8004), governed by a decentralized autonomous organization (DAO), and Google's Agent2Agent (A2A), governed by a corporate consortium, through a Sustainability by Design (SbD) lens. Using an LLM-powered pipeline combining automated annotation, neural topic modeling, and multi-layer network analysis, we identify contrasting governance and innovation architectures. ERC-8004 relies on permissionless participation, rough consensus, and decoupled deployment, while A2A assigns binding authority to an eight-seat Technology Steering Committee. The DAO concentrates on constitutive questions of trust and security, including what to build and why, whereas the consortium distributes attention across executive engineering questions of how to implement, document, and deliver the protocol. Both show high participation inequality, while corporate contributors span roughly twice as many themes as DAO contributors. We ask how these architectures produce distinct SDG-relevant signatures and what design principles they suggest for sustainable AI governance. We interpret institutional, discursive, and network patterns through SDGs 8, 9, 10, 11, 12, 16, and 17, identifying capacities for transparency, participation, contestability, and cross-protocol coordination. We argue that sustainable AI infrastructure requires a corrective feedback loop between designed charters and governance in practice, advancing SDG 16 on strong institutions. By integrating computational evidence, organizational research, and sustainable development, this review derives actionable design principles for sustainable AI governance.