{"paper":{"title":"Verifiable Agentic Infrastructure: Proof-Derived Authorization for Sovereign AI Systems","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Proof objects derived from consensus replace standing credentials to authorize actions by autonomous AI agents.","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"Deying Yu, Jun He","submitted_at":"2026-05-13T17:58:52Z","abstract_excerpt":"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. This risk becomes especially acute in sovereign AI systems, where autonomous agents may interact with cloud infrastructure, regulated data, financial workflows, and national-scale digital services. Governed mutation substrates reduce this risk "},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Under stated substrate assumptions, this architecture enforces a compact authorization invariant: no high-stakes execution without a proof object, no derived authority without consensus, and no valid mutation detached from evidence.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The framework depends on the existence and correct functioning of a governed mutation substrate capable of reliably interposing on agent actions and evaluating context and policy, as stated in the abstract.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Introduces DTF, a proof-derived authorization architecture using justification proofs, consensus evaluation, ephemeral identities, and evidence chains to govern AI agent actions on mutation substrates.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Proof objects derived from consensus replace standing credentials to authorize actions by autonomous AI agents.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"4593677fea896b79bb5e985a04d81eed390c1cc1a368d8fb2bd529cf545c8e8b"},"source":{"id":"2605.15228","kind":"arxiv","version":1},"verdict":{"id":"ea06f89d-dc35-4625-9004-eb801c02234d","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T17:26:00.517050Z","strongest_claim":"Under stated substrate assumptions, this architecture enforces a compact authorization invariant: no high-stakes execution without a proof object, no derived authority without consensus, and no valid mutation detached from evidence.","one_line_summary":"Introduces DTF, a proof-derived authorization architecture using justification proofs, consensus evaluation, ephemeral identities, and evidence chains to govern AI agent actions on mutation substrates.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The framework depends on the existence and correct functioning of a governed mutation substrate capable of reliably interposing on agent actions and evaluating context and policy, as stated in the abstract.","pith_extraction_headline":"Proof objects derived from consensus replace standing credentials to authorize actions by autonomous AI agents."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.15228/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"claim_evidence","ran_at":"2026-05-19T18:01:56.106819Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T17:36:29.116227Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_title_agreement","ran_at":"2026-05-19T17:31:18.491285Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T13:33:22.831479Z","status":"skipped","version":"1.0.0","findings_count":0}],"snapshot_sha256":"e5ff3c2c5189403baf08c7d83c83453be0060539d9a5a12dd377bc2dc4aac17c"},"references":{"count":17,"sample":[{"doi":"","year":2026,"title":"OpenKedge: Governing Agentic Mutation with Execution-Bound Safety and Evidence Chains","work_id":"5bbc8b0c-2fe4-4585-b9ce-199ef293678a","ref_index":1,"cited_arxiv_id":"2604.08601","is_internal_anchor":true},{"doi":"","year":1992,"title":"Ravi S. Sandhu, Edward J. Coyne, Hal L. Feinstein, and Charles E. Youman. Role-based access control models. InProceedings of the 15th National Computer Security Conference, 1992","work_id":"5c66ea23-6aad-4bc0-b402-50dfa46dd5d4","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2015,"title":"Hu, David Ferraiolo, Rick Kuhn, Arthur R","work_id":"9971f8dd-c7ce-4211-9a49-3e7611c88f1c","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2020,"title":"Zero trust architecture","work_id":"14d9d56c-60a9-472b-bc45-40d283b210ed","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2024,"title":"Cedar: A new lan- guage for expressive, fast, safe, and analyzable authorization.Proceedings of the ACM on Programming Languages, 8(OOPSLA1):670–697, 2024","work_id":"b03b8b49-fe69-4c49-8eb7-20fa8865a18e","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":17,"snapshot_sha256":"63e4f6549b62834d3f140225f7963401dcaf8169ac65e17ab15804a5befddcef","internal_anchors":2},"formal_canon":{"evidence_count":2,"snapshot_sha256":"4b4b8bd789365b05b2822bb546ef55a76a4b6053393426ba235db02a73f14aad"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}