AI is doing more of the modeling behind a forecast, but the authority and responsibility for what goes to the board doesn't move with it. This panel gets into what it actually means to hold that line: how CFOs are making sure models are trained on data they'd actually defend, what a real audit trail looks like so a forecast isn't a black box, and what happens when a number misses and "the model said so" isn't an acceptable answer. Key takeaways: - Why authority over the forecast stays with the CFO regardless of how much of the modeling AI handles. - What "properly trained" actually means in practice, and how CFOs are vetting the data before it ever reaches the model. - How CFOs are building a genuine audit trail into AI-generated forecasts, so the logic is inspectable, not a black box, when the board asks questions.