{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:TVK7MKASSB5ISYLQ66CFMJOZTR","short_pith_number":"pith:TVK7MKAS","schema_version":"1.0","canonical_sha256":"9d55f62812907a896170f7845625d99c46dbecd60da3149325c5c713e4950b34","source":{"kind":"arxiv","id":"2508.16481","version":3},"attestation_state":"computed","paper":{"title":"Benchmarking the Robustness of Agentic Systems to Adversarially-Induced Harms","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Adish Singla, Goran Radanovic, Jonathan N\\\"other","submitted_at":"2025-08-22T15:53:22Z","abstract_excerpt":"Ensuring the safe use of agentic systems requires a thorough understanding of the range of malicious behaviors these systems may exhibit. In this paper, we evaluate the robustness of LLM-based agentic systems against attacks that aim to elicit harmful actions from agents. To this end, we propose a novel taxonomy of harms for agentic systems and a novel benchmark, BAD-ACTS, for studying the security of agentic systems with respect to a wide range of harmful actions. BAD-ACTS consists of five implementations of agentic systems in distinct application environments, as well as a dataset of 238 hig"},"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":"2508.16481","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-08-22T15:53:22Z","cross_cats_sorted":[],"title_canon_sha256":"807bca6e92b48ac09e896e4521722e6ec44b7d50087e165f2c38cf8c9e2ff1fe","abstract_canon_sha256":"9f91b8b9974107c7456571ea96f547f3425dc9f332a3694d6a0be4001d7ebdfc"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-11T00:13:59.632152Z","signature_b64":"dPTR8r7COGrpBvhvr+fIvStKLOvHsT/yhtKXorn2oPHaxx0voMH3WMmvT4UQoS8GIevVTen3qAu/mjdAZjIGAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9d55f62812907a896170f7845625d99c46dbecd60da3149325c5c713e4950b34","last_reissued_at":"2026-08-11T00:13:59.629795Z","signature_status":"signed_v1","first_computed_at":"2026-08-11T00:13:59.629795Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Benchmarking the Robustness of Agentic Systems to Adversarially-Induced Harms","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Adish Singla, Goran Radanovic, Jonathan N\\\"other","submitted_at":"2025-08-22T15:53:22Z","abstract_excerpt":"Ensuring the safe use of agentic systems requires a thorough understanding of the range of malicious behaviors these systems may exhibit. In this paper, we evaluate the robustness of LLM-based agentic systems against attacks that aim to elicit harmful actions from agents. To this end, we propose a novel taxonomy of harms for agentic systems and a novel benchmark, BAD-ACTS, for studying the security of agentic systems with respect to a wide range of harmful actions. BAD-ACTS consists of five implementations of agentic systems in distinct application environments, as well as a dataset of 238 hig"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.16481","kind":"arxiv","version":3},"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/2508.16481/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":"2508.16481","created_at":"2026-08-11T00:13:59.630237+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.16481v3","created_at":"2026-08-11T00:13:59.630237+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.16481","created_at":"2026-08-11T00:13:59.630237+00:00"},{"alias_kind":"pith_short_12","alias_value":"TVK7MKASSB5I","created_at":"2026-08-11T00:13:59.630237+00:00"},{"alias_kind":"pith_short_16","alias_value":"TVK7MKASSB5ISYLQ","created_at":"2026-08-11T00:13:59.630237+00:00"},{"alias_kind":"pith_short_8","alias_value":"TVK7MKAS","created_at":"2026-08-11T00:13:59.630237+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":5,"sample":[{"citing_arxiv_id":"2606.19826","citing_title":"Heterogeneous LLM Debate Under Adversarial Peers: Honest Gains, Replacement Costs, and Resilience","ref_index":19,"is_internal_anchor":true},{"citing_arxiv_id":"2606.10749","citing_title":"Toward Secure LLM Agents: Threat Surfaces, Attacks, Defenses, and Evaluation","ref_index":134,"is_internal_anchor":true},{"citing_arxiv_id":"2605.16282","citing_title":"Taxonomy and Consistency Analysis of Safety Benchmarks for AI Agents","ref_index":41,"is_internal_anchor":true},{"citing_arxiv_id":"2605.18672","citing_title":"Position: A Three-Layer Probabilistic Assume-Guarantee Architecture Is Structurally Required for Safe LLM Agent Deployment","ref_index":43,"is_internal_anchor":true},{"citing_arxiv_id":"2604.23338","citing_title":"A Systematic Survey of Security Threats and Defenses in LLM-Based AI Agents: A Layered Attack Surface Framework","ref_index":143,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TVK7MKASSB5ISYLQ66CFMJOZTR","json":"https://pith.science/pith/TVK7MKASSB5ISYLQ66CFMJOZTR.json","graph_json":"https://pith.science/api/pith-number/TVK7MKASSB5ISYLQ66CFMJOZTR/graph.json","events_json":"https://pith.science/api/pith-number/TVK7MKASSB5ISYLQ66CFMJOZTR/events.json","paper":"https://pith.science/paper/TVK7MKAS"},"agent_actions":{"view_html":"https://pith.science/pith/TVK7MKASSB5ISYLQ66CFMJOZTR","download_json":"https://pith.science/pith/TVK7MKASSB5ISYLQ66CFMJOZTR.json","view_paper":"https://pith.science/paper/TVK7MKAS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.16481&json=true","fetch_graph":"https://pith.science/api/pith-number/TVK7MKASSB5ISYLQ66CFMJOZTR/graph.json","fetch_events":"https://pith.science/api/pith-number/TVK7MKASSB5ISYLQ66CFMJOZTR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TVK7MKASSB5ISYLQ66CFMJOZTR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TVK7MKASSB5ISYLQ66CFMJOZTR/action/storage_attestation","attest_author":"https://pith.science/pith/TVK7MKASSB5ISYLQ66CFMJOZTR/action/author_attestation","sign_citation":"https://pith.science/pith/TVK7MKASSB5ISYLQ66CFMJOZTR/action/citation_signature","submit_replication":"https://pith.science/pith/TVK7MKASSB5ISYLQ66CFMJOZTR/action/replication_record"}},"created_at":"2026-08-11T00:13:59.630237+00:00","updated_at":"2026-08-11T00:13:59.630237+00:00"}