{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:TYV4YA3T7JIEQL6FLI64EYOZ3X","short_pith_number":"pith:TYV4YA3T","schema_version":"1.0","canonical_sha256":"9e2bcc0373fa50482fc55a3dc261d9ddd2beed2b7383c4234486ff634000633e","source":{"kind":"arxiv","id":"2607.07695","version":1},"attestation_state":"computed","paper":{"title":"Institutional Red-Teaming: Deployment Rules, Not Just Models, Causally Shape Multi-Agent AI Safety","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.GT","cs.MA"],"primary_cat":"cs.AI","authors_text":"Yujiao Chen","submitted_at":"2026-07-08T17:53:56Z","abstract_excerpt":"We introduce institutional red-teaming, an evaluation methodology for testing deployment rules in multi-agent AI: hold the agents, objectives, and task state fixed, vary only one rule, and attribute the resulting change in collective behavior to that rule. We instantiate the methodology in IABench-CA, a consequence-allocation benchmark spanning 228 contexts, five canonical rules, and seven model populations (33,924 games), with a normative cooperative reference and auto-labelled reasoning traces. Three findings emerge. (1) Deployment rules causally alter collective safety: changing only the co"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2607.07695","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-07-08T17:53:56Z","cross_cats_sorted":["cs.GT","cs.MA"],"title_canon_sha256":"792a9f57f9b1e305fb610e9288d21cf0f7add77f496af5c4f332ec6a6ec66275","abstract_canon_sha256":"26a716c83e7d11f89f5a300cdf95a7b77b94138522924da3c75d75f85b2adc19"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-09T01:20:39.375257Z","signature_b64":"oygMnV7csI5Ty/z+Wr3VDJ5LGet5BWTnHGkmuvjgNUXysSVWeSSISdbqRodVDIH61C+Ne0rvS/1+EZcyoOABCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9e2bcc0373fa50482fc55a3dc261d9ddd2beed2b7383c4234486ff634000633e","last_reissued_at":"2026-07-09T01:20:39.374840Z","signature_status":"signed_v1","first_computed_at":"2026-07-09T01:20:39.374840Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Institutional Red-Teaming: Deployment Rules, Not Just Models, Causally Shape Multi-Agent AI Safety","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.GT","cs.MA"],"primary_cat":"cs.AI","authors_text":"Yujiao Chen","submitted_at":"2026-07-08T17:53:56Z","abstract_excerpt":"We introduce institutional red-teaming, an evaluation methodology for testing deployment rules in multi-agent AI: hold the agents, objectives, and task state fixed, vary only one rule, and attribute the resulting change in collective behavior to that rule. We instantiate the methodology in IABench-CA, a consequence-allocation benchmark spanning 228 contexts, five canonical rules, and seven model populations (33,924 games), with a normative cooperative reference and auto-labelled reasoning traces. Three findings emerge. (1) Deployment rules causally alter collective safety: changing only the co"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.07695","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2607.07695/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2607.07695","created_at":"2026-07-09T01:20:39.374905+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.07695v1","created_at":"2026-07-09T01:20:39.374905+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.07695","created_at":"2026-07-09T01:20:39.374905+00:00"},{"alias_kind":"pith_short_12","alias_value":"TYV4YA3T7JIE","created_at":"2026-07-09T01:20:39.374905+00:00"},{"alias_kind":"pith_short_16","alias_value":"TYV4YA3T7JIEQL6F","created_at":"2026-07-09T01:20:39.374905+00:00"},{"alias_kind":"pith_short_8","alias_value":"TYV4YA3T","created_at":"2026-07-09T01:20:39.374905+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TYV4YA3T7JIEQL6FLI64EYOZ3X","json":"https://pith.science/pith/TYV4YA3T7JIEQL6FLI64EYOZ3X.json","graph_json":"https://pith.science/api/pith-number/TYV4YA3T7JIEQL6FLI64EYOZ3X/graph.json","events_json":"https://pith.science/api/pith-number/TYV4YA3T7JIEQL6FLI64EYOZ3X/events.json","paper":"https://pith.science/paper/TYV4YA3T"},"agent_actions":{"view_html":"https://pith.science/pith/TYV4YA3T7JIEQL6FLI64EYOZ3X","download_json":"https://pith.science/pith/TYV4YA3T7JIEQL6FLI64EYOZ3X.json","view_paper":"https://pith.science/paper/TYV4YA3T","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.07695&json=true","fetch_graph":"https://pith.science/api/pith-number/TYV4YA3T7JIEQL6FLI64EYOZ3X/graph.json","fetch_events":"https://pith.science/api/pith-number/TYV4YA3T7JIEQL6FLI64EYOZ3X/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TYV4YA3T7JIEQL6FLI64EYOZ3X/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TYV4YA3T7JIEQL6FLI64EYOZ3X/action/storage_attestation","attest_author":"https://pith.science/pith/TYV4YA3T7JIEQL6FLI64EYOZ3X/action/author_attestation","sign_citation":"https://pith.science/pith/TYV4YA3T7JIEQL6FLI64EYOZ3X/action/citation_signature","submit_replication":"https://pith.science/pith/TYV4YA3T7JIEQL6FLI64EYOZ3X/action/replication_record"}},"created_at":"2026-07-09T01:20:39.374905+00:00","updated_at":"2026-07-09T01:20:39.374905+00:00"}