{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:HFJYPCLFMJQUYGLSXP3EOG3BWD","short_pith_number":"pith:HFJYPCLF","schema_version":"1.0","canonical_sha256":"395387896562614c1972bbf6471b61b0c3e6b8295c9da1767f62075805b40101","source":{"kind":"arxiv","id":"2303.11156","version":4},"attestation_state":"computed","paper":{"title":"Can AI-Generated Text be Reliably Detected?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Aounon Kumar, Soheil Feizi, Sriram Balasubramanian, Vinu Sankar Sadasivan, Wenxiao Wang","submitted_at":"2023-03-17T17:53:19Z","abstract_excerpt":"Large Language Models (LLMs) perform impressively well in various applications. However, the potential for misuse of these models in activities such as plagiarism, generating fake news, and spamming has raised concern about their responsible use. Consequently, the reliable detection of AI-generated text has become a critical area of research. AI text detectors have shown to be effective under their specific settings. In this paper, we stress-test the robustness of these AI text detectors in the presence of an attacker. We introduce recursive paraphrasing attack to stress test a wide range of d"},"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":"2303.11156","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-03-17T17:53:19Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"47b3662d629f6f12b29204e70a810dfd8a3866f104c5b1a3111cb7c645e05d87","abstract_canon_sha256":"9ec02186dd810c50091cdf530dbbd5b0f0fd696e916b5e9030022d5334c42405"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T19:23:20.655025Z","signature_b64":"ZyznNWb8T+7qW7hRojmiQAPi45IaqW9l0dRjGNvjcQN1MHqfHMj2+eXPQKtDRgReOT32qRa1umlRIZfgRdkxCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"395387896562614c1972bbf6471b61b0c3e6b8295c9da1767f62075805b40101","last_reissued_at":"2026-05-20T19:23:20.653222Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T19:23:20.653222Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Can AI-Generated Text be Reliably Detected?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Aounon Kumar, Soheil Feizi, Sriram Balasubramanian, Vinu Sankar Sadasivan, Wenxiao Wang","submitted_at":"2023-03-17T17:53:19Z","abstract_excerpt":"Large Language Models (LLMs) perform impressively well in various applications. However, the potential for misuse of these models in activities such as plagiarism, generating fake news, and spamming has raised concern about their responsible use. Consequently, the reliable detection of AI-generated text has become a critical area of research. AI text detectors have shown to be effective under their specific settings. In this paper, we stress-test the robustness of these AI text detectors in the presence of an attacker. We introduce recursive paraphrasing attack to stress test a wide range of d"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2303.11156","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/2303.11156/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":"2303.11156","created_at":"2026-05-20T19:23:20.653332+00:00"},{"alias_kind":"arxiv_version","alias_value":"2303.11156v4","created_at":"2026-05-20T19:23:20.653332+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2303.11156","created_at":"2026-05-20T19:23:20.653332+00:00"},{"alias_kind":"pith_short_12","alias_value":"HFJYPCLFMJQU","created_at":"2026-05-20T19:23:20.653332+00:00"},{"alias_kind":"pith_short_16","alias_value":"HFJYPCLFMJQUYGLS","created_at":"2026-05-20T19:23:20.653332+00:00"},{"alias_kind":"pith_short_8","alias_value":"HFJYPCLF","created_at":"2026-05-20T19:23:20.653332+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":25,"internal_anchor_count":3,"sample":[{"citing_arxiv_id":"2605.16336","citing_title":"Detecting Verbatim LLM Copy-Paste in Homework","ref_index":11,"is_internal_anchor":true},{"citing_arxiv_id":"2605.16436","citing_title":"The End of Trust: How Agentic AI Breaks Security Assumptions","ref_index":51,"is_internal_anchor":true},{"citing_arxiv_id":"2303.11156","citing_title":"Can AI-Generated Text be Reliably Detected?","ref_index":62,"is_internal_anchor":true},{"citing_arxiv_id":"2605.16107","citing_title":"Multi-Level Contextual Token Relation Modeling for Machine-Generated Text Detection","ref_index":53,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16471","citing_title":"From AI-Generated Content to Agentic Action: Security and Safety Threats in Generative AI","ref_index":115,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16035","citing_title":"Who Owns This Agent? Tracing AI Agents Back to Their Owners","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19516","citing_title":"Base Models Look Human To AI Detectors","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19402","citing_title":"High-Rate Public-Key Pseudorandom Codes for Edit Errors","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19915","citing_title":"LLM Agents Make Collective Belief Dynamics Programmable: Challenges and Research Directions","ref_index":43,"is_internal_anchor":false},{"citing_arxiv_id":"2507.07871","citing_title":"Mitigating Watermark Forgery in Generative Models via Randomized Key Selection","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2509.22055","citing_title":"RedNote-Vibe: A Dataset for Capturing Temporal Dynamics of AI-Generated Text in Lifestyle Social Media","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2603.00177","citing_title":"Detecting Cognitive Signatures in Typing Behavior for Non-Intrusive Authorship Verification","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2603.00179","citing_title":"Privacy-Preserving Proof of Human Authorship via Zero-Knowledge Process Attestation","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2603.23146","citing_title":"Why AI-Generated Text Detection Fails: Evidence from Explainable AI Beyond Benchmark Accuracy","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12452","citing_title":"The Algorithmic Caricature: Auditing LLM-Generated Political Discourse Across Crisis Events","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2311.05232","citing_title":"A Survey on Hallucination in Large Language Models: Principles, Taxonomy, Challenges, and Open Questions","ref_index":278,"is_internal_anchor":false},{"citing_arxiv_id":"2605.03210","citing_title":"Human-Provenance Verification should be Treated as Labor Infrastructure in AI-Saturated Markets","ref_index":58,"is_internal_anchor":false},{"citing_arxiv_id":"2604.26328","citing_title":"DSIPA: Detecting LLM-Generated Texts via Sentiment-Invariant Patterns Divergence Analysis","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06524","citing_title":"Process Matters more than Output for Distinguishing Humans from Machines","ref_index":59,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06524","citing_title":"Process Matters more than Output for Distinguishing Humans from Machines","ref_index":59,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05503","citing_title":"Chainwash: Multi-Step Rewriting Attacks on Diffusion Language Model Watermarks","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2605.00348","citing_title":"Block-wise Codeword Embedding for Reliable Multi-bit Text Watermarking","ref_index":56,"is_internal_anchor":false},{"citing_arxiv_id":"2604.10893","citing_title":"Beyond A Fixed Seal: Adaptive Stealing Watermark in Large Language Models","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06342","citing_title":"\"Don't Be Afraid, Just Learn\": Insights from Industry Practitioners to Prepare Software Engineers in the Age of Generative AI","ref_index":52,"is_internal_anchor":false},{"citing_arxiv_id":"2604.14513","citing_title":"PeerPrism: Peer Evaluation Expertise vs Review-writing AI","ref_index":29,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HFJYPCLFMJQUYGLSXP3EOG3BWD","json":"https://pith.science/pith/HFJYPCLFMJQUYGLSXP3EOG3BWD.json","graph_json":"https://pith.science/api/pith-number/HFJYPCLFMJQUYGLSXP3EOG3BWD/graph.json","events_json":"https://pith.science/api/pith-number/HFJYPCLFMJQUYGLSXP3EOG3BWD/events.json","paper":"https://pith.science/paper/HFJYPCLF"},"agent_actions":{"view_html":"https://pith.science/pith/HFJYPCLFMJQUYGLSXP3EOG3BWD","download_json":"https://pith.science/pith/HFJYPCLFMJQUYGLSXP3EOG3BWD.json","view_paper":"https://pith.science/paper/HFJYPCLF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2303.11156&json=true","fetch_graph":"https://pith.science/api/pith-number/HFJYPCLFMJQUYGLSXP3EOG3BWD/graph.json","fetch_events":"https://pith.science/api/pith-number/HFJYPCLFMJQUYGLSXP3EOG3BWD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HFJYPCLFMJQUYGLSXP3EOG3BWD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HFJYPCLFMJQUYGLSXP3EOG3BWD/action/storage_attestation","attest_author":"https://pith.science/pith/HFJYPCLFMJQUYGLSXP3EOG3BWD/action/author_attestation","sign_citation":"https://pith.science/pith/HFJYPCLFMJQUYGLSXP3EOG3BWD/action/citation_signature","submit_replication":"https://pith.science/pith/HFJYPCLFMJQUYGLSXP3EOG3BWD/action/replication_record"}},"created_at":"2026-05-20T19:23:20.653332+00:00","updated_at":"2026-05-20T19:23:20.653332+00:00"}