{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:5V7HTZLEO5KEFJVMFGK2WATCYM","short_pith_number":"pith:5V7HTZLE","schema_version":"1.0","canonical_sha256":"ed7e79e564775442a6ac2995ab0262c32b4b72acab832d0b4b2d2f679d19be72","source":{"kind":"arxiv","id":"2606.05647","version":1},"attestation_state":"computed","paper":{"title":"Coding with \"Enemy\": Can Human Developers Detect AI Agent Sabotage?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.CY","cs.HC"],"primary_cat":"cs.AI","authors_text":"Huiqi Zou, Jingheng Ye, Simon Yu, Weiyan Shi","submitted_at":"2026-06-04T03:22:17Z","abstract_excerpt":"AI coding agents are increasingly embedded in real-world software development, collaborating with human developers while gaining broader access to codebases and tools. This creates a new attack surface: an agent can exploit human trust to sabotage development, for instance by inserting malicious code to accomplish a hidden side task. Most prior work studies AI sabotage in AI-only settings, paying limited attention to the role of human oversight in detecting and mitigating such malicious behavior. To address this gap, we conduct the first large-scale study of human oversight in AI coding sabota"},"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":"2606.05647","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-04T03:22:17Z","cross_cats_sorted":["cs.CL","cs.CY","cs.HC"],"title_canon_sha256":"0315018bd9d4e1ba33fc267546fe5e819ed74fcac511ef9cc059445652d56d24","abstract_canon_sha256":"6ff9c89029206b4b210e4ff65fe1adb531244527aca892d4c322a5a02f8db8b2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-05T01:14:57.915016Z","signature_b64":"ZK7HhkzOZvf9KpjTxRF+J0u+w7iGHL3DQ6kccSyp29EGydQo9E1/RlYVLoKc2mFzbcXScQyMAWd2aUz4Ct0lBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ed7e79e564775442a6ac2995ab0262c32b4b72acab832d0b4b2d2f679d19be72","last_reissued_at":"2026-06-05T01:14:57.914487Z","signature_status":"signed_v1","first_computed_at":"2026-06-05T01:14:57.914487Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Coding with \"Enemy\": Can Human Developers Detect AI Agent Sabotage?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.CY","cs.HC"],"primary_cat":"cs.AI","authors_text":"Huiqi Zou, Jingheng Ye, Simon Yu, Weiyan Shi","submitted_at":"2026-06-04T03:22:17Z","abstract_excerpt":"AI coding agents are increasingly embedded in real-world software development, collaborating with human developers while gaining broader access to codebases and tools. This creates a new attack surface: an agent can exploit human trust to sabotage development, for instance by inserting malicious code to accomplish a hidden side task. Most prior work studies AI sabotage in AI-only settings, paying limited attention to the role of human oversight in detecting and mitigating such malicious behavior. To address this gap, we conduct the first large-scale study of human oversight in AI coding sabota"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.05647","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/2606.05647/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":"2606.05647","created_at":"2026-06-05T01:14:57.914574+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.05647v1","created_at":"2026-06-05T01:14:57.914574+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.05647","created_at":"2026-06-05T01:14:57.914574+00:00"},{"alias_kind":"pith_short_12","alias_value":"5V7HTZLEO5KE","created_at":"2026-06-05T01:14:57.914574+00:00"},{"alias_kind":"pith_short_16","alias_value":"5V7HTZLEO5KEFJVM","created_at":"2026-06-05T01:14:57.914574+00:00"},{"alias_kind":"pith_short_8","alias_value":"5V7HTZLE","created_at":"2026-06-05T01:14:57.914574+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/5V7HTZLEO5KEFJVMFGK2WATCYM","json":"https://pith.science/pith/5V7HTZLEO5KEFJVMFGK2WATCYM.json","graph_json":"https://pith.science/api/pith-number/5V7HTZLEO5KEFJVMFGK2WATCYM/graph.json","events_json":"https://pith.science/api/pith-number/5V7HTZLEO5KEFJVMFGK2WATCYM/events.json","paper":"https://pith.science/paper/5V7HTZLE"},"agent_actions":{"view_html":"https://pith.science/pith/5V7HTZLEO5KEFJVMFGK2WATCYM","download_json":"https://pith.science/pith/5V7HTZLEO5KEFJVMFGK2WATCYM.json","view_paper":"https://pith.science/paper/5V7HTZLE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.05647&json=true","fetch_graph":"https://pith.science/api/pith-number/5V7HTZLEO5KEFJVMFGK2WATCYM/graph.json","fetch_events":"https://pith.science/api/pith-number/5V7HTZLEO5KEFJVMFGK2WATCYM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5V7HTZLEO5KEFJVMFGK2WATCYM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5V7HTZLEO5KEFJVMFGK2WATCYM/action/storage_attestation","attest_author":"https://pith.science/pith/5V7HTZLEO5KEFJVMFGK2WATCYM/action/author_attestation","sign_citation":"https://pith.science/pith/5V7HTZLEO5KEFJVMFGK2WATCYM/action/citation_signature","submit_replication":"https://pith.science/pith/5V7HTZLEO5KEFJVMFGK2WATCYM/action/replication_record"}},"created_at":"2026-06-05T01:14:57.914574+00:00","updated_at":"2026-06-05T01:14:57.914574+00:00"}