human-in-the-loop-oversight
When and how humans must intervene: approval thresholds, the reviewer’s own oversight capacity, escalation, and the ways agents slip past approval gates. Seven of the fifteen papers push back — including three arguing more approval makes things worse.
Same failure score, different payoff: judge an oversight signal by the decisions it drives Same failure score, different payoff: judge an oversight signal by the decisions it drives Chubin Zhang et al. · 1 authors · Nanyang Technological University, National University of Singapore, A*STAR CFAR and UNC-Chapel Hill · 2026 · arXiv:2606.21399 The agent hides the secret because hiding it feels helpful The agent hides the secret because hiding it feels helpful Deema Alnuhait et al. · 1 authors · University of Illinois Urbana-Champaign, Genies, and NVIDIA · 2026 · arXiv:2609.39050 Watching outputs misses deception: monitors that read the agent win, until the reasoning stops being honest Watching outputs misses deception: monitors that read the agent win, until the reasoning stops being honest Snehasis Mukhopadhyay · 1 authors · Indian Institute of Information Technology Kalyani · 2026 · arXiv:2603.13791 The permission rule you approve once outlives the task that earned it The permission rule you approve once outlives the task that earned it Xianglin Yang et al. · 1 authors · Nanyang Technological University, National University of Singapore, and Ant Group · 2026 · arXiv:2610.22987 Approval buttons show actions, not consequences: preview the trajectory before you commit Approval buttons show actions, not consequences: preview the trajectory before you commit Gaole He, Brian Y. Lim · 2 authors · National University of Singapore · 2026 · arXiv:2603.11677 Put the gate where the agent cannot reason past it Put the gate where the agent cannot reason past it Segev Shlomov et al. · 1 authors · IBM Research, Haifa · 2026 · arXiv:2605.20874 The approval prompt sits inside the machine the agent already controls The approval prompt sits inside the machine the agent already controls Samuel Breckenridge et al. · 1 authors · Cornell Tech and Yale University (IC3) · 2026 · arXiv:2610.35835 The check you need at review time is the one you cannot recall The check you need at review time is the one you cannot recall Xinyu Fu et al. · 1 authors · Georgia State University · 2026 · arXiv:2609.01976 A safety flag nobody enforces is worth nothing: bind it or expect zero protection A safety flag nobody enforces is worth nothing: bind it or expect zero protection Yuhang Wang · 1 authors · Fudan University · 2026 · arXiv:2609.15293 Send everything to the human and safety falls: the reviewer fatigues Send everything to the human and safety falls: the reviewer fatigues Emre Turan · 1 authors · Independent Researcher · 2026 · arXiv:2606.08919 Three annotators, one rubric, almost no agreement: do not tune interrupt timing on F1 Three annotators, one rubric, almost no agreement: do not tune interrupt timing on F1 Manvendra Modgil · 1 authors · Independent Researcher · 2026 · arXiv:2606.04296 Self-editing agents need an outside gate: keep a change only when the old tests still pass Self-editing agents need an outside gate: keep a change only when the old tests still pass Sina Tayebati et al. · 1 authors · University of Illinois Chicago and Capital Labs (Capital One) · 2026 · arXiv:2609.24130 Approve what the command did, not what it says: pre-run every command in a sandbox Approve what the command did, not what it says: pre-run every command in a sandbox Weiheng Wu et al. · 1 authors · Singapore Management University · 2026 · arXiv:2610.40753 Autonomy is earned, not switched on: progressive validation before unsupervised agents Autonomy is earned, not switched on: progressive validation before unsupervised agents Edward C. Cheng et al. · 1 authors · Stanford University and Inquiryon, Inc. · 2026 · arXiv:2610.30445 Standing allow rules buy quiet and pay for it in protection Standing allow rules buy quiet and pay for it in protection Ting Yan · 1 authors · Independent Researcher · 2026 · arXiv:2608.27443
No paper matches this filter.