Purpose & Intent
Every AI Product must declare its purpose, intended use, and distinct intelligence proposition.
Purpose defines why the Product exists. Intent defines the boundaries within which it is intended and approved to operate. The intelligence proposition states the specific artificial-intelligence capability being productized and is therefore part of AI Product qualification.
Purpose Declaration
An AI Product must declare:
- Problem Statement — the challenge it addresses;
- Intended Outcomes — the value it is expected to create;
- Target Domain — relevant business or operating context;
- Business Value — economic, operational, societal, or user benefit.
Purpose is a contextual dimension and must not be used as AI Product Type.
Distinct Intelligence Proposition
The Product must state an intelligence proposition that can be independently understood and governed.
Examples:
- detect payment fraud;
- classify customer intent;
- reason over a regulated knowledge base;
- recommend suitable actions;
- plan autonomous navigation;
- generate governed product imagery.
A statement such as “uses an LLM” is not an intelligence proposition. It describes an implementation dependency rather than the value-bearing intelligence being productized.
Intent Declaration
An AI Product must define intended usage boundaries:
- Supported Use Cases — validated tasks/workflows;
- Consumer Scope — intended consumer populations and organizational scope;
- Constraints — operating assumptions and prerequisites;
- Exclusions — explicit out-of-scope uses;
- Human / automated decision boundary — where relevant to Agency / Autonomy.
Relationship to AI Qualification
Purpose alone does not establish AI Product Family membership.
The declared intelligence proposition contributes to AI Product Qualification & Evidence, which also requires:
- BPS Product qualification;
- intelligence as the value-bearing essence;
- a qualifying AI Primitive path;
- a governed Product boundary.
Example
AI Product: Fraud Detection Intelligence
- Problem Statement: Identify transactions exhibiting characteristics materially associated with fraud.
- Intelligence Proposition: Detect and classify payment fraud risk with evidence suitable for downstream review.
- Target Domain: Financial services / payments.
- Supported Uses: Transaction scoring, alerting, fraud investigation prioritization.
- Consumer Scope: Risk and payment operations.
- Constraints: Requires governed transaction features and registered input contracts.
- Exclusions: Not intended as a standalone AML or sanctions decision capability.
Summary
- Purpose describes why the Product exists.
- Intent describes where and how it should be used.
- The intelligence proposition describes what artificial intelligence is actually being productized.
Principle: A credible AI Product must state the intelligence proposition it owns, not merely the AI technology it uses.