---
id: "action-redefine-spans-of-control"
type: "action-item"
source_timestamps: ["§ 1. Explicitly redefine workflows and then name new (human) role expectations."]
tags: ["org-design", "management"]
related: ["concept-oversight-capacity", "concept-ai-brain-fry"]
action: "Redesign team sizes and managerial spans of control to account for the cognitive limits of overseeing high-volume AI output."
outcome: "Prevents 'AI brain fry' and maintains high quality control by ensuring humans have adequate bandwidth to review automated work."
sources: ["agentic"]
sourceVaultSlug: "hbr-seg-agentic"
originDay: 6
articleStem: "hbr-ext-16-dont-treat-agents-like-employees"
sourceUrl: "https://hbr.org/2026/05/research-why-you-shouldnt-treat-ai-agents-like-employees"
sourceTitle: "Research: Why You Shouldn’t Treat AI Agents Like Employees"
---
# Redefine Spans of Control for AI Oversight

**Action:** Redesign team sizes and managerial spans of control to account for the cognitive limits of overseeing high-volume AI output.

**Expected outcome:** Prevents [[concept-ai-brain-fry]] and maintains high quality control by ensuring humans have adequate bandwidth to review automated work.

This is the operational core of Step 1 of the [[framework-responsible-human-ai-collaboration]]. It follows directly from [[concept-oversight-capacity]] and the principle in [[quote-oversight-capacity]] that oversight capacity does not scale with output. Skipping it produces the error spikes measured in [[claim-brain-fry-errors]] and [[claim-quality-control-decline]]. Pair with [[action-reset-performance-management]] so oversight quality is actually rewarded. Requires the prerequisite [[prereq-org-design-basics]].