Definition
AI governance is the set of structures through which an organization decides how AI may be used, who is accountable for its outputs, and how that accountability is evidenced: policy, risk classification, human oversight, escalation paths, record-keeping and review. It is distinct from AI ethics, which sets the principles, and from AI compliance, which demonstrates conformity with a particular regulation.
Warum es wichtig ist
Most AI governance programmes stop at the policy document. The policy states that a human reviews high-impact recommendations; nothing in the architecture records whether a human actually did, what they saw, or whether they disagreed. Governance that cannot be evidenced is a statement of intent, and it fails at exactly the moment it is needed. This is also why the ownership question is so persistent: risk, legal, data and engineering each hold a piece, and governance breaks wherever no one owns the record itself.
Woher es kommt
An established discipline drawing on corporate governance, model risk management and emerging AI regulation. The Human Layer series contributes the architectural half of the argument: oversight has to be built into systems, because oversight that depends on intention degrades into rubber-stamping.
Lesen Sie die zugrunde liegenden Papers: The Human Layer · The Human Layer Audit