liveBook · Manning

Strategy and policy level

As explained in AI Governance

The 6L-G governance level that establishes organizational direction, AI policies, leadership, accountability, and oversight.

Strategy and Policy is the top level of the 6L-G governance model. It sets organizational direction and defines boundaries, leadership, accountability, and oversight before teams make AI-specific design or deployment choices. section 1.3

Make the organization’s AI rules operational by defining:

  • Which AI uses are prohibited, which require formal approval, and who may approve exceptions. section 3.2
  • Approved tools, named owners, human-oversight requirements, and review paths for high-risk initiatives. section 3.2 section 1.3
  • A governance charter, responsible-AI principles, executive sponsorship, and legal and ethical boundaries. section 1.3

MediAssist patient chatbot

Before a chatbot speaks directly to patients, reads medical records, or routes urgent cases to clinicians or emergency services, policy should require review and a named owner. The policy sets these conditions; it does not specify exactly when emergency services should be contacted. section 3.2

Without concrete policy, teams may use unauthorized tools with customer data or launch sensitive AI features without review. Clinical rules, tests, logs, and handoff procedures belong to later controls. section 3.2

Figure 1.4 Six Levels of Generative AI Governance. Chapter 2 will expand into what control tasks attach to each checkpoint.

Use the six-level model to locate Strategy and Policy among the governance checkpoints tied to an AI system’s evolution. section 1.3

How AI Governance covers it

How deep it goes
Defines approval, ownership, and human-oversight expectations, but leaves case-specific clinical rules, tests, logs, and handoffs to later controls.
Examples
Uses MediAssist patient escalation and unauthorized chatbot use with customer data to illustrate policy gaps.
What you can do after
Configure AI governance by defining prohibited uses, approved tools, exception authority, review requirements, named owners, and human-oversight conditions.
How it connects
Connects to the 6L-G model, Responsible AI principles, governance charters, risk assessment, and the PDCA cycle.

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.