Skip to content

Tool

Human approval planner

Describe one AI task and see the level of human control it needs, with a control pattern your team can set up.

About the task

Can the action be undone?

Who sees the output?

Does it touch money, legal terms or safety?

How many items does it handle a week?

How much harm does an occasional error do?

Can the system show its confidence or its source?

Settings your team sets
0/6answered

How the scoring works

  • A professional sign-off, such as legal advice or a credit decision, stays manual, with AI limited to gathering and summarizing inputs.
  • An action that touches money, legal terms or safety directly, or that cannot be undone, needs a named person to approve every item.
  • Otherwise the answers add risk points: up to 2 automates fully, 3 to 5 samples, and 6 or more needs approval on every item.
  • An irreversible or high-stakes action where errors do serious harm stays manual if the system cannot show its confidence or source, because a reviewer would have nothing to check against.
  • The sample share, confidence threshold and escalation time are settings your team chooses. The starting values are placeholders, not recommendations.
  • The level is a starting point for planning. Test it on real past cases before anything goes live, and review it as results come in.
Read the full guide

Go deeper

One useful email a month.

New guides, tools and practical notes on putting AI to work. No sales sequences.

Planning where people stay in control?

Bring us the task. We will help you design the approval points before anything is built.

Book a discovery call