Applicable (Part 4): ATTRIBUTE 1 - Applicable

A team produces a deliverable that an AI quality gate assesses before it advances to the next stage. The human is the Producer. The AI is the Accepter. This is already happening at scale in software development, legal review support, content moderation, and compliance checking — and it is growing rapidly as organizations discover that AI can apply acceptance criteria faster, more consistently, and at greater volume than human reviewers alone. This work type raises a governance question that none of the other five surface directly: when is AI acceptance sufficient, and when must a human Accepter remain in the loop? WEEFAI’s answer is structural rather than prescriptive. The Definer must specify not only what is being produced but what level of acceptance authority the AI holds — whether AI evaluation is the final gate, a preliminary filter before human review, or an advisory signal the human Accepter weighs alongside their own judgment. The stakes of the output determine the answer. A security scan that flags a critical vulnerability requires human review before any decision is made. A grammar and style check on internal documentation may reasonably be AI-final. The framework does not collapse these distinctions. It requires that they be made explicitly, in advance, by the Definer — which is precisely where that accountability belongs. These six work types — AI Assisted, AI Augmented, AI Generated, Non-AI Human Work Outcomes, Agentic AI, and AI Evaluated — cover the full spectrum of how knowledge work is now produced, delivered, and reviewed. A Planner sequencing a real workweek will encounter all six. The framework does not require separate systems for each. It requires a single system that can see all of them clearly and govern each appropriately. That is what Applicability means in practice: not that the framework bends to fit every context, but that it was built to hold all of them at once.