In most energy companies AI governance sits wherever the first AI tool was installed, usually IT. That placement is the problem: IT can govern the tools, but only the board and executive leadership can own the decisions AI now shapes. What NIST says about ownership, who should own which layer, and five signs governance is in the wrong place.
Machine learning already sits inside the software that produces a reserves estimate. Regulation S-K Item 1202(a)(7) already requires you to describe the internal controls over that effort. Six questions that bring a stale control narrative back in line with the process you actually run.
Defensible AI has nothing to do with how advanced your models are. It is whether you can reconstruct how an AI-influenced decision was made, who reviewed it, and on whose authority. A six-item checklist for energy operators.
