{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:AWEE55QYUPPESEDDLW5LLYXFKG","short_pith_number":"pith:AWEE55QY","schema_version":"1.0","canonical_sha256":"05884ef618a3de4910635dbab5e2e551a218fd208f196e44b2f7abfb9e2749ad","source":{"kind":"arxiv","id":"2409.16427","version":4},"attestation_state":"computed","paper":{"title":"HAICOSYSTEM: An Ecosystem for Sandboxing Safety Risks in Human-AI Interactions","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Bill Yuchen Lin, Faeze Brahman, Frank Xu, Hao Zhu, Hyunwoo Kim, Liwei Jiang, Maarten Sap, Niloofar Mireshghallah, Ronan Le Bras, Ximing Lu, Xuhui Zhou, Yejin Choi","submitted_at":"2024-09-24T19:47:21Z","abstract_excerpt":"AI agents are increasingly autonomous in their interactions with human users and tools, leading to increased interactional safety risks. We present HAICOSYSTEM, a framework examining AI agent safety within diverse and complex social interactions. HAICOSYSTEM features a modular sandbox environment that simulates multi-turn interactions between human users and AI agents, where the AI agents are equipped with a variety of tools (e.g., patient management platforms) to navigate diverse scenarios (e.g., a user attempting to access other patients' profiles). To examine the safety of AI agents in thes"},"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":"2409.16427","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2024-09-24T19:47:21Z","cross_cats_sorted":[],"title_canon_sha256":"846136eca9dc5075704f8e0d55a252a76058e09f70ff185d86e90c72c94f4e13","abstract_canon_sha256":"d989ad3b6d8dd1fe3acf10d889957addc2a55c34f46dc99ba4b1ccff4053243d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:01:54.864968Z","signature_b64":"qAR0vN3qGSSVT4/tUHi+1QCe73Wp/wLnOsToIO2Is/s8ufFj93jP8dPzIVBSy9xDBSIEzyHE8TGOy97GgNOnBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"05884ef618a3de4910635dbab5e2e551a218fd208f196e44b2f7abfb9e2749ad","last_reissued_at":"2026-07-05T12:01:54.864479Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:01:54.864479Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"HAICOSYSTEM: An Ecosystem for Sandboxing Safety Risks in Human-AI Interactions","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Bill Yuchen Lin, Faeze Brahman, Frank Xu, Hao Zhu, Hyunwoo Kim, Liwei Jiang, Maarten Sap, Niloofar Mireshghallah, Ronan Le Bras, Ximing Lu, Xuhui Zhou, Yejin Choi","submitted_at":"2024-09-24T19:47:21Z","abstract_excerpt":"AI agents are increasingly autonomous in their interactions with human users and tools, leading to increased interactional safety risks. We present HAICOSYSTEM, a framework examining AI agent safety within diverse and complex social interactions. HAICOSYSTEM features a modular sandbox environment that simulates multi-turn interactions between human users and AI agents, where the AI agents are equipped with a variety of tools (e.g., patient management platforms) to navigate diverse scenarios (e.g., a user attempting to access other patients' profiles). To examine the safety of AI agents in thes"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.16427","kind":"arxiv","version":4},"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/2409.16427/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":"2409.16427","created_at":"2026-07-05T12:01:54.864533+00:00"},{"alias_kind":"arxiv_version","alias_value":"2409.16427v4","created_at":"2026-07-05T12:01:54.864533+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.16427","created_at":"2026-07-05T12:01:54.864533+00:00"},{"alias_kind":"pith_short_12","alias_value":"AWEE55QYUPPE","created_at":"2026-07-05T12:01:54.864533+00:00"},{"alias_kind":"pith_short_16","alias_value":"AWEE55QYUPPESEDD","created_at":"2026-07-05T12:01:54.864533+00:00"},{"alias_kind":"pith_short_8","alias_value":"AWEE55QY","created_at":"2026-07-05T12:01:54.864533+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.20506","citing_title":"Reinforcing Human Behavior Simulation via Verbal Feedback","ref_index":61,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09721","citing_title":"Security Risks in Tool-Enabled AI Agents: A Systematic Analysis of Privileged Execution Environments","ref_index":21,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/AWEE55QYUPPESEDDLW5LLYXFKG","json":"https://pith.science/pith/AWEE55QYUPPESEDDLW5LLYXFKG.json","graph_json":"https://pith.science/api/pith-number/AWEE55QYUPPESEDDLW5LLYXFKG/graph.json","events_json":"https://pith.science/api/pith-number/AWEE55QYUPPESEDDLW5LLYXFKG/events.json","paper":"https://pith.science/paper/AWEE55QY"},"agent_actions":{"view_html":"https://pith.science/pith/AWEE55QYUPPESEDDLW5LLYXFKG","download_json":"https://pith.science/pith/AWEE55QYUPPESEDDLW5LLYXFKG.json","view_paper":"https://pith.science/paper/AWEE55QY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2409.16427&json=true","fetch_graph":"https://pith.science/api/pith-number/AWEE55QYUPPESEDDLW5LLYXFKG/graph.json","fetch_events":"https://pith.science/api/pith-number/AWEE55QYUPPESEDDLW5LLYXFKG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AWEE55QYUPPESEDDLW5LLYXFKG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AWEE55QYUPPESEDDLW5LLYXFKG/action/storage_attestation","attest_author":"https://pith.science/pith/AWEE55QYUPPESEDDLW5LLYXFKG/action/author_attestation","sign_citation":"https://pith.science/pith/AWEE55QYUPPESEDDLW5LLYXFKG/action/citation_signature","submit_replication":"https://pith.science/pith/AWEE55QYUPPESEDDLW5LLYXFKG/action/replication_record"}},"created_at":"2026-07-05T12:01:54.864533+00:00","updated_at":"2026-07-05T12:01:54.864533+00:00"}