{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:2CIEN7ACGKWKY36LIK4COE3N2G","short_pith_number":"pith:2CIEN7AC","schema_version":"1.0","canonical_sha256":"d09046fc0232acac6fcb42b827136dd197efa86f14577f949a5c7dab8ac7b3bf","source":{"kind":"arxiv","id":"2508.21634","version":1},"attestation_state":"computed","paper":{"title":"Human-Written vs. AI-Generated Code: A Large-Scale Study of Defects, Vulnerabilities, and Complexity","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.SE","authors_text":"Cristina Improta, Domenico Cotroneo, Pietro Liguori","submitted_at":"2025-08-29T13:51:28Z","abstract_excerpt":"As AI code assistants become increasingly integrated into software development workflows, understanding how their code compares to human-written programs is critical for ensuring reliability, maintainability, and security. In this paper, we present a large-scale comparison of code authored by human developers and three state-of-the-art LLMs, i.e., ChatGPT, DeepSeek-Coder, and Qwen-Coder, on multiple dimensions of software quality: code defects, security vulnerabilities, and structural complexity. Our evaluation spans over 500k code samples in two widely used languages, Python and Java, classif"},"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":"2508.21634","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2025-08-29T13:51:28Z","cross_cats_sorted":[],"title_canon_sha256":"ec8d93b2c4c97fe21994b1836d8a4a1f89f1a356cfe7d1b06d406537598e9ab7","abstract_canon_sha256":"b0cf2ddda8f795152b9ba9fb7e131193a50ece09875743c6bc67d0f78ecfa7bc"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:01:46.996068Z","signature_b64":"J7HkQnjgvX8sGYGBuhrQhdb6dL1tZLl30lxBHawoswup1cbHNXUU1/A2QRVVsxkAqnbPn88mNG+bKTZ2+RLwBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d09046fc0232acac6fcb42b827136dd197efa86f14577f949a5c7dab8ac7b3bf","last_reissued_at":"2026-07-05T12:01:46.995559Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:01:46.995559Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Human-Written vs. AI-Generated Code: A Large-Scale Study of Defects, Vulnerabilities, and Complexity","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.SE","authors_text":"Cristina Improta, Domenico Cotroneo, Pietro Liguori","submitted_at":"2025-08-29T13:51:28Z","abstract_excerpt":"As AI code assistants become increasingly integrated into software development workflows, understanding how their code compares to human-written programs is critical for ensuring reliability, maintainability, and security. In this paper, we present a large-scale comparison of code authored by human developers and three state-of-the-art LLMs, i.e., ChatGPT, DeepSeek-Coder, and Qwen-Coder, on multiple dimensions of software quality: code defects, security vulnerabilities, and structural complexity. Our evaluation spans over 500k code samples in two widely used languages, Python and Java, classif"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.21634","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/2508.21634/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":"2508.21634","created_at":"2026-07-05T12:01:46.995618+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.21634v1","created_at":"2026-07-05T12:01:46.995618+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.21634","created_at":"2026-07-05T12:01:46.995618+00:00"},{"alias_kind":"pith_short_12","alias_value":"2CIEN7ACGKWK","created_at":"2026-07-05T12:01:46.995618+00:00"},{"alias_kind":"pith_short_16","alias_value":"2CIEN7ACGKWKY36L","created_at":"2026-07-05T12:01:46.995618+00:00"},{"alias_kind":"pith_short_8","alias_value":"2CIEN7AC","created_at":"2026-07-05T12:01:46.995618+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.17675","citing_title":"Bridging the Gap on AI-Assisted Scientific Software Development Through Transparency and Traceability","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06464","citing_title":"To What Extent Does Agent-generated Code Require Maintenance? An Empirical Study","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06464","citing_title":"To What Extent Does Agent-generated Code Require Maintenance? An Empirical Study","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2604.19965","citing_title":"Insights into Security-Related AI-Generated Pull Requests","ref_index":76,"is_internal_anchor":false},{"citing_arxiv_id":"2604.18334","citing_title":"Reliability of AI Bots Footprints in GitHub Actions CI/CD Workflows","ref_index":9,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/2CIEN7ACGKWKY36LIK4COE3N2G","json":"https://pith.science/pith/2CIEN7ACGKWKY36LIK4COE3N2G.json","graph_json":"https://pith.science/api/pith-number/2CIEN7ACGKWKY36LIK4COE3N2G/graph.json","events_json":"https://pith.science/api/pith-number/2CIEN7ACGKWKY36LIK4COE3N2G/events.json","paper":"https://pith.science/paper/2CIEN7AC"},"agent_actions":{"view_html":"https://pith.science/pith/2CIEN7ACGKWKY36LIK4COE3N2G","download_json":"https://pith.science/pith/2CIEN7ACGKWKY36LIK4COE3N2G.json","view_paper":"https://pith.science/paper/2CIEN7AC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.21634&json=true","fetch_graph":"https://pith.science/api/pith-number/2CIEN7ACGKWKY36LIK4COE3N2G/graph.json","fetch_events":"https://pith.science/api/pith-number/2CIEN7ACGKWKY36LIK4COE3N2G/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2CIEN7ACGKWKY36LIK4COE3N2G/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2CIEN7ACGKWKY36LIK4COE3N2G/action/storage_attestation","attest_author":"https://pith.science/pith/2CIEN7ACGKWKY36LIK4COE3N2G/action/author_attestation","sign_citation":"https://pith.science/pith/2CIEN7ACGKWKY36LIK4COE3N2G/action/citation_signature","submit_replication":"https://pith.science/pith/2CIEN7ACGKWKY36LIK4COE3N2G/action/replication_record"}},"created_at":"2026-07-05T12:01:46.995618+00:00","updated_at":"2026-07-05T12:01:46.995618+00:00"}