AI Governance Definition & Accountability Framework A plan to establish a unified definition of AI governance grounded in decision authority, accountability, and enforceable oversight mechanisms. Submitter's Note: This StratML rendition has been inferred by ChatGPT from Basil C. Puglisi's article "AI Governance Has No Formal Definition. That Is a Problem for Every Organization Using AI." It has been lightly edited in the form at https://stratml.us/forms/Claude/Part1.html AI Governance Ecosystem AIGE 8b7b0f85-eb33-49e3-9a80-9b952e3ba71e Institutions, organizations, and individuals responsible for defining, implementing, and enforcing AI governance. Basil C. Puglisi Author proposing a formal definition of AI governance grounded in accountability and decision authority. Standards Organizations Entities such as ISO responsible for governance and management system standards. Regulators Authorities establishing binding obligations for AI systems. Organizations Enterprises implementing AI systems and governance structures. Accountable Individuals Named humans holding decision authority and accountability for AI outputs. Clearly defined AI governance enabling accountable human oversight of AI systems across all institutional contexts. df31dc18-dbfb-4775-a7be-16ff38d0dec4 To establish a unified definition of AI governance grounded in decision authority, accountability, and enforceable oversight mechanisms. 4e10a3c7-d7aa-416a-938e-cbfe296a2c42 Accountability Ensure identifiable individuals are responsible for AI decisions and outcomes. Authority Assign decision-making power to qualified individuals with the ability to approve, modify, or halt AI outputs. Transparency Enable traceability and auditability of decisions affecting AI outputs. Responsibility Align AI use with ethical, legal, and societal expectations. Clarity Define governance concepts unambiguously across institutional contexts. Definition Establish a clear and unified definition of AI governance across institutions and frameworks. 4d647729-eef9-44e2-985d-fd068ba8acf9 1 None of the major institutional frameworks define AI governance explicitly. Conceptual Definition Define AI governance as decision authority, accountability structures, and oversight mechanisms exercised by named humans. eab849ab-a045-45e4-af62-f2dfdd39e798 1.1 Applied Definition Specify governance as the presence of a qualified human with binding authority and accountability at defined checkpoints. 34775da4-ea95-4e3d-9c03-132b92680cd3 1.2 Framework Alignment Integrate governance concepts across institutional frameworks and standards. 6e6578d5-c685-45f2-897a-1ab137e0784e 2 ISO, NIST, OECD, EU AI Act, and other frameworks provide partial but non-integrated perspectives. Accountability Structures Incorporate accountability and decision authority as defined in organizational governance standards. 3efc5de0-ae88-4b2d-85a3-487603dc46c8 2.1 Human Oversight Ensure alignment with principles requiring human responsibility and oversight of AI systems. 6fb599f4-a9b5-445e-8eca-2198aed0c123 2.2 Risk Posture Embed governance within organizational culture and risk management practices. b42a02d7-5faf-4fe6-a59d-abdca21e3e20 2.3 Accountability Ensure governance includes enforceable accountability for AI outputs. c32aa603-32c6-4126-b335-95a04a1f3ba5 3 Governance requires a named human whose decisions can be audited and held accountable. Decision Authority Assign binding authority to individuals to approve, modify, or halt AI outputs. 8ae77216-11b2-4cb7-bb06-cd04a2569166 3.1 Auditability Enable tracing and review of decisions affecting AI outputs. 9e0a3e87-bbca-48e4-8ebb-913e238939a6 3.2 Liability Channels Ensure accountability through moral, employment, civil, and criminal mechanisms. eaef4394-8816-4133-94b4-bdd58cd05192 3.3 Distinction Differentiate AI governance from responsible AI and management systems. 5fd7ece3-68b8-4950-a5ad-5f90a077c7cd 4 Responsible AI focuses on engineering; governance requires accountable human authority. Responsible AI Define responsible AI as engineering practices embedding ethical principles into systems. 4ada199d-f0d5-4652-850f-28fb751b19ec 4.1 Governance Threshold Establish the requirement of named human accountability as the threshold for governance. 68896da0-e240-463a-8024-3301c4dc3922 4.2 Operational Test Provide criteria for determining whether AI governance exists in practice. 8e5dd74f-bc24-4ee6-acb6-66f6cb9ddadc 5 Named Individual Identify a specific person with authority over AI outputs. 595d80c1-6d8d-48cb-8c38-be22af7d240e 5.1 Traceability Ensure decisions can be documented and reviewed. eb46db45-4091-4eea-94a3-67035ec2b5da 5.2 Accountability Channels Confirm applicability of moral, employment, civil, or criminal accountability. 8ac55d05-ec1f-499e-8444-e224d192be24 5.3 2026-03-15 2026-03-18 https://basilpuglisi.com/ai-governance-has-no-formal-definition-here-is-one/ Owen Ambur Owen.Ambur@verizon.net