Verifiable Agentic Infrastructure: Proof-Derived Authorization for Sovereign AI Systems
The paper introduces a Distributed Trust Framework (DTF) designed to enhance authorization for sovereign AI systems. It addresses the risks posed by autonomous AI agents that can perform unsafe actions despite having valid credentials. By shifting the focus from identity-based authorization to proof-derived authority, DTF aims to create a more governable and auditable infrastructure for AI deployments.
- ▪Modern systems rely on identity-centric authorization, which is challenged by autonomous AI agents.
- ▪The Distributed Trust Framework (DTF) provides a verification framework for governed mutation systems.
- ▪DTF enforces a compact authorization invariant that ensures no high-stakes execution occurs without a proof object.
Opening excerpt (first ~120 words) tap to expand
Computer Science > Artificial Intelligence arXiv:2605.15228 (cs) [Submitted on 13 May 2026] Title:Verifiable Agentic Infrastructure: Proof-Derived Authorization for Sovereign AI Systems Authors:Jun He, Deying Yu View a PDF of the paper titled Verifiable Agentic Infrastructure: Proof-Derived Authorization for Sovereign AI Systems, by Jun He and 1 other authors View PDF HTML (experimental) Abstract:Modern cloud and enterprise systems rely on identity-centric authorization, assuming that callers possessing valid credentials are safe to execute commands. The emergence of autonomous AI agents invalidates this assumption: agents can generate syntactically valid but semantically unsafe actions, making standing privileges a significant operational risk.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at arXiv cs.AI.
Discussion
0 commentsMore from arXiv cs.AI
