[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"glossary-approval-fatigue::en":3,"gloss-cluster-approval-fatigue::en":23,"gloss-next-approval-fatigue::en":9},{"slug":4,"category":5,"name":6,"definition":7,"meta_desc":8,"faq":9,"schema_markup":9,"related":10},"approval-fatigue","agents","Approval Fatigue","Approval fatigue is what happens to a human checkpoint at volume: the reviewer stops reading. It is the standard failure of oversight in automated systems, and it is a design problem rather than a discipline problem, because nobody sustains genuine scrutiny over a queue that never empties and is almost always fine.\n\nThe mechanism is worth stating plainly. A gate that fires on every action trains the person behind it that approving is the correct response, since it nearly always is. The rare case that deserved attention arrives looking exactly like the hundred before it, on the same screen, with the same button. Volume does not merely reduce the quality of review — past a certain rate it converts review into a formality that still appears in the process documentation and in the vendor's security questionnaire.\n\nThe symptom is easy to measure and rarely looked at. An approval rate close to a hundred per cent is not evidence that the system is working; it is evidence that the gate is either unnecessary or unread, and the two are distinguishable only by what happens on the cases that were wrong. Time spent per decision tells the same story: approvals that take a second or two are not decisions.\n\nThe fix is to gate less, not to exhort people to concentrate. Send to review only what is consequential — the irreversible, the unusual, the high-value, the low-confidence — and let the rest run with clear logging and a cheap undo. Sample the unreviewed stream rather than ignoring it. And give the reviewer what makes this case different from the ordinary one, since the useful judgement is comparative and the interface usually withholds exactly the comparison.\n\nThere is a second-order cost that shows up later. Once a team has learned that approvals are a formality, the habit persists into the cases that genuinely needed a decision, including in unrelated systems. Oversight is a limited budget, and spending it on cases that never needed it is how a company arrives at an incident review with a full audit trail of approvals that nobody remembers making.","Approval fatigue is a human checkpoint that stops being read at volume. An approval rate near a hundred per cent means the gate is unnecessary or unread.",null,[11,14,17,20],{"slug":12,"name":13},"approval-workflow","Approval Workflow",{"slug":15,"name":16},"confidence-threshold","Confidence Threshold",{"slug":18,"name":19},"escalation-path","Escalation Path",{"slug":21,"name":22},"human-in-the-loop","Human-in-the-Loop",[24,28,32,35,38,41,44,47,50,53,56,59],{"slug":25,"category":5,"name":26,"updated_at":27},"acceptance-criteria","Acceptance Criteria","2026-09-08T21:32:37+00:00",{"slug":29,"category":5,"name":30,"updated_at":31},"agent-budget","Agent Budget","2026-09-08T21:32:36+00:00",{"slug":33,"category":5,"name":34,"updated_at":31},"agent-checkpointing","Agent Checkpointing",{"slug":36,"category":5,"name":37,"updated_at":31},"agent-handoff","Agent Handoff",{"slug":39,"category":5,"name":40,"updated_at":27},"agent-identity","Agent Identity",{"slug":42,"category":5,"name":43,"updated_at":31},"agent-loop","Agent Loop",{"slug":45,"category":5,"name":46,"updated_at":31},"agent-memory","Agent Memory",{"slug":48,"category":5,"name":49,"updated_at":31},"agent-sandbox","Agent Sandbox",{"slug":51,"category":5,"name":52,"updated_at":27},"agent-simulation","Agent Simulation",{"slug":54,"category":5,"name":55,"updated_at":31},"agent-trajectory","Agent Trajectory",{"slug":57,"category":5,"name":58,"updated_at":31},"agentic-rag","Agentic RAG",{"slug":60,"category":5,"name":61,"updated_at":31},"computer-use","Computer Use"]