Is Decentralized AI Governable? From Regulative Policy to Constitutive Protocol
A new paper published on arXiv proposes a framework for governing decentralized AI (DeAI) systems. The authors identify a "governance vacuum" where DeAI systems are consequential but lack identifiable or controllable operators, leading to accountability and incapacitation gaps. They argue for a shift from policy-based regulation to protocol-based governance, embedding constraints within the system's architecture rather than relying on normative address to agents. AI
IMPACT Proposes a new governance model for decentralized AI, shifting focus from policy to protocol to address accountability gaps.