---
id: "action-define-decision-rights"
type: "action-item"
source_timestamps: ["§ 2. Make accountability explicit and personal for employees and agents."]
tags: ["governance", "risk-management"]
related: ["framework-accountability-rules", "concept-accountability-blurring"]
action: "Explicitly document what AI agents can do autonomously, what requires human approval, and the exact triggers for escalation."
outcome: "Eliminates accountability blurring by making it unequivocally clear which human is responsible for the final output and when they must intervene."
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"
---
# Define Human-AI Decision Rights and Escalation Paths

**Action:** Explicitly document what AI agents can do **autonomously**, what requires **human approval**, and the exact **triggers for escalation**.

**Expected outcome:** Eliminates [[concept-accountability-blurring]] by making it unequivocally clear which human is responsible for the final output and when they must intervene.

This operationalizes Step 2 of the [[framework-responsible-human-ai-collaboration]] and directly implements the three fronts of the [[framework-accountability-rules]] — decision rights, escalation, and consequences. It is the concrete antidote to the measured accountability shift in [[claim-accountability-shift-d6]] and the sentiment in [[quote-blame-technology]]. The unresolved legal backdrop is [[question-legal-accountability]].


## Related across articles
- [[action-design-hesitation]]
- [[action-govern-system]]
- [[concept-human-in-the-loop-escalation]]