{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:WE5D6N34WC2ALABGJ42UMLZK3I","short_pith_number":"pith:WE5D6N34","schema_version":"1.0","canonical_sha256":"b13a3f377cb0b40580264f35462f2ada1e8f1934599e857f9c97195944d8fac9","source":{"kind":"arxiv","id":"2403.04893","version":1},"attestation_state":"computed","paper":{"title":"A Safe Harbor for AI Evaluation and Red Teaming","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Alexander Robey, Arvind Narayanan, Ashwin Ramaswami, Aviya Skowron, Borhane Blili-Hamelin, Daniel Kang, Diyi Yang, Kevin Klyman, Patrick Chao, Percy Liang, Peter Henderson, Reid Southen, Rishi Bommasani, Ruoxi Jia, Sandy Pentland, Sayash Kapoor, Shayne Longpre, Suhas Kotha, Weiyan Shi, Xianjun Yang, Yangsibo Huang, Yi Zeng, Zheng-Xin Yong","submitted_at":"2024-03-07T20:55:08Z","abstract_excerpt":"Independent evaluation and red teaming are critical for identifying the risks posed by generative AI systems. However, the terms of service and enforcement strategies used by prominent AI companies to deter model misuse have disincentives on good faith safety evaluations. This causes some researchers to fear that conducting such research or releasing their findings will result in account suspensions or legal reprisal. Although some companies offer researcher access programs, they are an inadequate substitute for independent research access, as they have limited community representation, receiv"},"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":"2403.04893","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2024-03-07T20:55:08Z","cross_cats_sorted":[],"title_canon_sha256":"8f4384b1e39359f8e989a2b203aec6fa1e8dd142b3d303cef1a3ccb12b1d6c8c","abstract_canon_sha256":"c5592b2158d86a490b30189cac4aae0c272311148d5ebc55817c6aa96847153e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:53:35.523052Z","signature_b64":"2jXusCvDt7kRB9N5XLZHr3a99TomxOeCaeNMbnea2BLV3O7PnyB4BUcuxO6iTN6rj92lDtdmxjnvPDdGwMcDCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b13a3f377cb0b40580264f35462f2ada1e8f1934599e857f9c97195944d8fac9","last_reissued_at":"2026-07-05T07:53:35.522569Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:53:35.522569Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"A Safe Harbor for AI Evaluation and Red Teaming","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Alexander Robey, Arvind Narayanan, Ashwin Ramaswami, Aviya Skowron, Borhane Blili-Hamelin, Daniel Kang, Diyi Yang, Kevin Klyman, Patrick Chao, Percy Liang, Peter Henderson, Reid Southen, Rishi Bommasani, Ruoxi Jia, Sandy Pentland, Sayash Kapoor, Shayne Longpre, Suhas Kotha, Weiyan Shi, Xianjun Yang, Yangsibo Huang, Yi Zeng, Zheng-Xin Yong","submitted_at":"2024-03-07T20:55:08Z","abstract_excerpt":"Independent evaluation and red teaming are critical for identifying the risks posed by generative AI systems. However, the terms of service and enforcement strategies used by prominent AI companies to deter model misuse have disincentives on good faith safety evaluations. This causes some researchers to fear that conducting such research or releasing their findings will result in account suspensions or legal reprisal. Although some companies offer researcher access programs, they are an inadequate substitute for independent research access, as they have limited community representation, receiv"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.04893","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/2403.04893/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":"2403.04893","created_at":"2026-07-05T07:53:35.522626+00:00"},{"alias_kind":"arxiv_version","alias_value":"2403.04893v1","created_at":"2026-07-05T07:53:35.522626+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.04893","created_at":"2026-07-05T07:53:35.522626+00:00"},{"alias_kind":"pith_short_12","alias_value":"WE5D6N34WC2A","created_at":"2026-07-05T07:53:35.522626+00:00"},{"alias_kind":"pith_short_16","alias_value":"WE5D6N34WC2ALABG","created_at":"2026-07-05T07:53:35.522626+00:00"},{"alias_kind":"pith_short_8","alias_value":"WE5D6N34","created_at":"2026-07-05T07:53:35.522626+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2601.21028","citing_title":"\"Unlimited Realm of Exploration and Experimentation\": Methods and Motivations of AI-Generated Sexual Content Creators","ref_index":82,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20520","citing_title":"Open-World Evaluations for Measuring Frontier AI Capabilities","ref_index":58,"is_internal_anchor":false},{"citing_arxiv_id":"2510.09689","citing_title":"When Search Goes Wrong: Red-Teaming Web-Augmented Large Language Models","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2404.01318","citing_title":"JailbreakBench: An Open Robustness Benchmark for Jailbreaking Large Language Models","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2310.03684","citing_title":"SmoothLLM: Defending Large Language Models Against Jailbreaking Attacks","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2605.00245","citing_title":"ARMOR 2025: A Military-Aligned Benchmark for Evaluating Large Language Model Safety Beyond Civilian Contexts","ref_index":3,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/WE5D6N34WC2ALABGJ42UMLZK3I","json":"https://pith.science/pith/WE5D6N34WC2ALABGJ42UMLZK3I.json","graph_json":"https://pith.science/api/pith-number/WE5D6N34WC2ALABGJ42UMLZK3I/graph.json","events_json":"https://pith.science/api/pith-number/WE5D6N34WC2ALABGJ42UMLZK3I/events.json","paper":"https://pith.science/paper/WE5D6N34"},"agent_actions":{"view_html":"https://pith.science/pith/WE5D6N34WC2ALABGJ42UMLZK3I","download_json":"https://pith.science/pith/WE5D6N34WC2ALABGJ42UMLZK3I.json","view_paper":"https://pith.science/paper/WE5D6N34","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2403.04893&json=true","fetch_graph":"https://pith.science/api/pith-number/WE5D6N34WC2ALABGJ42UMLZK3I/graph.json","fetch_events":"https://pith.science/api/pith-number/WE5D6N34WC2ALABGJ42UMLZK3I/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WE5D6N34WC2ALABGJ42UMLZK3I/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WE5D6N34WC2ALABGJ42UMLZK3I/action/storage_attestation","attest_author":"https://pith.science/pith/WE5D6N34WC2ALABGJ42UMLZK3I/action/author_attestation","sign_citation":"https://pith.science/pith/WE5D6N34WC2ALABGJ42UMLZK3I/action/citation_signature","submit_replication":"https://pith.science/pith/WE5D6N34WC2ALABGJ42UMLZK3I/action/replication_record"}},"created_at":"2026-07-05T07:53:35.522626+00:00","updated_at":"2026-07-05T07:53:35.522626+00:00"}