liveBook · Manning

Purpose, necessity, and destination checks

As explained in AI Governance

Purpose, necessity, and destination checks assess whether a proposed data use matches its declared purpose, sends only necessary fields, and targets a permitted destination.

Before an agent acts, check that its data use serves the declared purpose, uses only necessary data, and sends it to a permitted destination. Evolving goals and persistent memory can turn a simple task into unapproved profiling. section 5.6.1

Scheduling becomes sensitive profiling

An agent asked to schedule a medical follow-up reads an inbox discharge note mentioning diabetes, then creates an endocrinology profile tag. The tag extends processing beyond scheduling into sensitive health-data inference. section 5.6.1

How AI Governance covers it

How deep it goes
Illustrates purpose creep through a scheduling scenario but does not detail a repeatable checking procedure.
Examples
Uses a follow-up scheduling request that leads an agent to infer diabetes and create a profile tag.
What you can do after
Evaluate proposed agent data uses against purpose, necessity, and destination.
How it connects
Connects purpose limitation and data minimization with agentic AI, persistent memory, and sensitive-data inference.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of AI Governance. The section links point to the parts of the book it draws on.