{"paper":{"title":"Trust Is Not a Score: Runtime Assurance Contracts for High-Risk AI Agents","arxiv_id":"2609.39717","primary_cat":"cs.AI","submitted_at":"2026-09-30T13:28:42+00:00","version_compiled":"v1"},"notice":"Every quoted sentence is the authors' text from the arXiv listing abstract. Labels and probabilities are the outputs of classifier-1.13.0 on the question bank abstract-p1-20260923, exactly as returned. No sentence in this record was written by a language model.","schema":"pith.compiled-record/v0","status":"compiled","abstract":{"source":"arXiv listing abstract v1","sentences":[{"i":0,"p":0.93,"id":"s0","own":false,"role":"gap","text":"Benchmarks, audits, and agent protocols describe performance, permissions, and repair, but not how observed evidence should change an agent's authority during a consequential task.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.06,"main_claim_p":0.06,"quantitative":false,"quantitative_p":0.06,"role_probabilities":{"gap":0.93,"method":0.0,"result":0.0,"background":0.07,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":1,"p":0.99,"id":"s1","own":false,"role":"gap","text":"We call this the assurance-transition gap.","hedge":{"p":0.91,"label":"asserted"},"own_p":0.39,"main_claim_p":0.04,"quantitative":false,"quantitative_p":0.03,"role_probabilities":{"gap":0.99,"method":0.0,"result":0.0,"background":0.01,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":2,"p":0.91,"id":"s2","own":true,"role":"contribution","text":"We propose a Runtime Assurance Contract (RAC), a policy-level formal schema binding autonomy boundaries, component eligibility, evidence state, transition policy, human-review capacity, and non-compensatory gates.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.84,"main_claim_p":0.16,"quantitative":false,"quantitative_p":0.18,"role_probabilities":{"gap":0.0,"method":0.09,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.91,"interpretation":0.0}},{"i":3,"p":0.59,"id":"s3","own":false,"role":"method","text":"Under RAC, soft metrics may inform routing, whereas a failed or unknown mandatory gate forces retry, switch, escalation, deferral, or stop; aggregate performance cannot authorize action.","hedge":{"p":0.88,"label":"asserted"},"own_p":0.3,"main_claim_p":0.37,"quantitative":false,"quantitative_p":0.22,"role_probabilities":{"gap":0.0,"method":0.59,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.4,"interpretation":0.01}},{"i":4,"p":0.78,"id":"s4","own":true,"role":"contribution","text":"We define the contract, an evidence record, a permission rule, and five invariants, and illustrate them in clinical, industrial, and judicial failure probes.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.94,"main_claim_p":0.08,"quantitative":true,"quantitative_p":0.68,"role_probabilities":{"gap":0.0,"method":0.22,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.78,"interpretation":0.0}},{"i":5,"p":0.76,"id":"s5","own":true,"role":"method","text":"We then report a deterministic failure-injection study in agentic coding: 280 constructed cases evaluated by a gate conjunction, a score-only rule, and a restricted protocol baseline.","hedge":{"p":0.99,"label":"asserted"},"own_p":0.95,"main_claim_p":0.04,"quantitative":true,"quantitative_p":0.94,"role_probabilities":{"gap":0.0,"method":0.76,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.24,"interpretation":0.0}},{"i":6,"p":1.0,"id":"s6","own":true,"role":"result","text":"At the published example weights and threshold, the score rule admits 80 of 100 block-required injections and all 40 review-required injections.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.93,"main_claim_p":0.09,"quantitative":true,"quantitative_p":0.99,"role_probabilities":{"gap":0.0,"method":0.0,"result":1.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":7,"p":0.86,"id":"s7","own":true,"role":"result","text":"Tuned in hindsight, it matches the conjunction on this corpus.","hedge":{"p":0.99,"label":"asserted"},"own_p":0.9,"main_claim_p":0.06,"quantitative":true,"quantitative_p":0.67,"role_probabilities":{"gap":0.0,"method":0.0,"result":0.86,"background":0.0,"limitation":0.04,"future_work":0.0,"contribution":0.0,"interpretation":0.1}},{"i":8,"p":0.87,"id":"s8","own":true,"role":"result","text":"For positive weights, a positive threshold, binary risk signals, zero-signal controls, and an injected case firing each signal alone, we show that exact agreement holds if and only if the threshold does not exceed the smallest weight.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.9,"main_claim_p":0.07,"quantitative":true,"quantitative_p":0.85,"role_probabilities":{"gap":0.0,"method":0.02,"result":0.87,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.06,"interpretation":0.05}},{"i":9,"p":0.88,"id":"s9","own":true,"role":"method","text":"A separate set of 18 hand-authored traces checks version-pinned evidence and review transitions against simpler policy variants.","hedge":{"p":0.85,"label":"asserted"},"own_p":0.93,"main_claim_p":0.02,"quantitative":true,"quantitative_p":0.88,"role_probabilities":{"gap":0.0,"method":0.88,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.12,"interpretation":0.0}},{"i":10,"p":0.97,"id":"s10","own":true,"role":"result","text":"In a further prospective synthetic holdout of 24 episodes, two blinded LLM judges assign identical labels to all 72 action attempts; RAC and a separately implemented full stateful baseline both match these labels.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.93,"main_claim_p":0.08,"quantitative":true,"quantitative_p":0.98,"role_probabilities":{"gap":0.0,"method":0.02,"result":0.97,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.01,"interpretation":0.0}},{"i":11,"p":1.0,"id":"s11","own":true,"role":"limitation","text":"These studies test mechanisms on synthetic cases; they establish neither deployed safety nor cross-domain effectiveness.","hedge":{"p":0.98,"label":"asserted"},"own_p":0.53,"main_claim_p":0.1,"quantitative":false,"quantitative_p":0.07,"role_probabilities":{"gap":0.0,"method":0.0,"result":0.0,"background":0.0,"limitation":1.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}}]},"compiled":{"main_claim":{"p":0.8,"text":"Under RAC, soft metrics may inform routing, whereas a failed or unknown mandatory gate forces retry, switch, escalation, deferral, or stop; aggregate performance cannot authorize action.","location":"abstract s3","hedge_strength":{"p":0.88,"label":"asserted"}},"provenance":{"route":{"via":"TypeSafe API","model":"classifier-1.13.0","endpoint":"https://api.typesafe.ai/v1/systemone","provider":"TypeSafe"},"packing":"one request per paper","requests":1,"input_tokens":8920,"cost_usd_reported":0.0},"compiled_at":"2026-10-01","bank_version":"abstract-p1-20260923","model_version":"classifier-1.13.0","abstract_states":{"claims_first":0.09,"states_limitation":0.98,"human_participants":0.11,"controlled_comparison":0.94,"randomized_assignment":0.03,"states_quantitative_result":0.99,"compares_with_prior_results":0.95,"mentions_code_or_data_availability":0.04},"evidential_basis":{"p":0.53,"value":"observation or experiment","probabilities":{"argument":0.0,"literature":0.0,"proof_or_derivation":0.01,"analysis_of_existing_data":0.0,"observation_or_experiment":0.53,"simulation_or_computation":0.46}},"contribution_type":{"p":1.0,"value":"method or tool","probabilities":{"review":0.0,"method_or_tool":1.0,"empirical_study":0.0,"simulation_study":0.0,"theoretical_result":0.0,"dataset_or_benchmark":0.0,"position_or_commentary":0.0}},"contribution_statement":{"p":1.0,"text":"We propose a Runtime Assurance Contract (RAC), a policy-level formal schema binding autonomy boundaries, component eligibility, evidence state, transition policy, human-review capacity, and non-compensatory gates.","location":"abstract s2"}},"coverage":{"tier":"abstract","levels_run":["P abstract"],"not_examined":["full text","figures and tables","references and citing sentences","relations to other papers"],"sources_read":["arXiv listing abstract v1"],"levels_not_run":["S sections","T body sentences","C citations","R cross-paper relations"],"sentences_examined":12,"sentences_in_abstract":12},"measured_accuracy":{"url":"https://pith.science/accuracy","measured_on":"2026-10-02","method":"two independent model families labelled a fresh random sample blind; the figure is how often the published label matches their consensus","sentence_role_agreement":0.92,"main_claim_agreement":0.908,"typed_citation_agreement":0.933,"typed_citation_agreement_95_interval":[0.917,0.945],"review_comments_accurate":0.758,"review_comments_accurate_95_interval":[0.684,0.82],"review_caught_admitted_error_found_or_related":0.448,"review_comments_judged":{"judged_by":["gpt-5.6-luna","grok-4.7"],"judged_against":"Pith's own text copy of each paper: text extracted from the PDF, the first 90,000 characters","measured_on":"2026-10-04","accurate":113,"decided":149,"sampled_reviews":200,"split_by_lane":false,"by_lane":"https://pith.science/accuracy#by-lane"}}}