{"paper":{"title":"Some[Body] Must Receive That Pain for Agent Accountability","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"AI agents cause harm but lack any persistent body to receive consequences and change behavior, so high-stakes use must stay tethered to human principals.","cross_cats":["cs.AI"],"primary_cat":"cs.CY","authors_text":"Botao Amber Hu, Helena Rong","submitted_at":"2026-05-16T08:24:06Z","abstract_excerpt":"AI agents increasingly act consequentially in the real world. This creates a problem we call \\emph{consequence reception}: harm occurs, the producing system is identified, yet no continuing agent receives consequences in a way that changes future behavior. Pain, understood mechanistically as a corrective feedback signal, is foundational to canonical theories of punishment -- deterrence, rehabilitation, retribution, and incapacitation all assume a continuing locus that registers the signal and updates behavior. That, in turn, requires a body for the signal to land on: a boundary whose integrity"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Achieving consequence-agency coupling is therefore a sociotechnical infrastructural problem, not only a legal one. Until such architectures exist, high-stakes AI deployment should remain tethered to accountable human principals with meaningful control, proportional liability, and authority to constrain or terminate the agent.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The assumption that pain, understood mechanistically as a corrective feedback signal, is foundational to canonical theories of punishment and requires a continuing locus (body) for effective consequence reception in AI agents.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"AI agents lack the persistent identity and feedback mechanisms needed for consequence reception, requiring new architectures or continued human accountability.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"AI agents cause harm but lack any persistent body to receive consequences and change behavior, so high-stakes use must stay tethered to human principals.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"e98f1bd31d9b6029b9eef972541bd33f684f3b6449559df7f68b3f1695e3b97d"},"source":{"id":"2605.16872","kind":"arxiv","version":1},"verdict":{"id":"a17cc122-2c87-462c-8f02-d9da143a20fe","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T19:42:06.127127Z","strongest_claim":"Achieving consequence-agency coupling is therefore a sociotechnical infrastructural problem, not only a legal one. Until such architectures exist, high-stakes AI deployment should remain tethered to accountable human principals with meaningful control, proportional liability, and authority to constrain or terminate the agent.","one_line_summary":"AI agents lack the persistent identity and feedback mechanisms needed for consequence reception, requiring new architectures or continued human accountability.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The assumption that pain, understood mechanistically as a corrective feedback signal, is foundational to canonical theories of punishment and requires a continuing locus (body) for effective consequence reception in AI agents.","pith_extraction_headline":"AI agents cause harm but lack any persistent body to receive consequences and change behavior, so high-stakes use must stay tethered to human principals."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.16872/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_title_agreement","ran_at":"2026-05-19T20:01:18.974980Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T19:50:42.528341Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"claim_evidence","ran_at":"2026-05-19T18:41:56.297228Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T18:33:26.373747Z","status":"skipped","version":"1.0.0","findings_count":0}],"snapshot_sha256":"e9e5c7b0f6dd2a74d68a45948898841338791b72e1dcc7195a2f54db0eabe642"},"references":{"count":107,"sample":[{"doi":"","year":2017,"title":"Artificial Intelligence and Law , volume=","work_id":"f9bc21a9-de72-48b4-a10e-e8ad422f36df","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2018,"title":"2019 , month = nov, url =","work_id":"131e7f3e-a352-40e2-aba6-ade66ea0987e","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2019,"title":"UC Davis Law Review , volume=","work_id":"7978f856-4403-419f-a782-6aabc049e31b","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2018,"title":"Nature Human Behaviour , volume=","work_id":"34bce5e3-019d-43b4-85b8-e83915ebcc6f","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2004,"title":"Psychological Science , volume=","work_id":"5d4fb7ad-ff2a-49cf-b637-540b40ff2318","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":107,"snapshot_sha256":"9f795d902000014497131741fe485bfb56926dcf03ca148aee8acceb9d8809d2","internal_anchors":8},"formal_canon":{"evidence_count":2,"snapshot_sha256":"8a18972975fa2d80a2dd786558f349108bea0d23ced92d7515e9e5ea7a8a4580"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}