NIST AI Risk Management Framework
Primary guidance on human oversight, accountability, and managing consequential AI risks.
← Back to Compute College
Compute College
Human approval and reversible actions
Design approval gates and undo paths before an AI system can take consequential actions.
Plain-English definition
Human approval is a deliberate checkpoint before an AI-generated action crosses a consequential boundary. Reversible actions can be undone or safely retried, while irreversible actions require stronger authorization, preview, audit, and escalation controls.
Memory trick: Preview before impact; reverse what you can.
Why it matters
A model can produce a plausible request without understanding authority, side effects, or the user’s intent. Approval and reversibility limit blast radius when model, tool, or context errors occur.
Simple example
An assistant may draft an email and show recipients, attachments, and text for approval. It may create a reversible calendar hold, but sending a contract or deleting a record requires a separate authenticated confirmation.
Example figures are illustrative calculations, not current quoted market prices.
Current example
Source material
Common mistake
A vague “run” button gives a human nothing useful to approve. The review must show the exact action, inputs, target, side effects, and identity under which it will execute.
Practical takeaway
What you can do with this
Classify actions by impact and reversibility. Add preview, dry run, approval, idempotency, audit, and rollback behavior before exposing each tool.
Decision check: can a reviewer understand and undo the exact action before it happens?
Compute College learning path
AI Engineering
Step 34 of 48: Human approval and reversible actions