AI is already inside most finance functions, from analysts pasting spreadsheets into chatbots to vendors promising instant answers. Boards want speed, but the CFO is the one who has to defend the numbers. This session explores what separates AI that impresses from AI you can trust: governed business context, humans in the loop where it matters, and a fix for the messy middle of fragmented data. You’ll leave with a clearer way to judge any AI approach and see it applied in a live planning cycle, from forecast to close. Key takeaways: - Trust as a leadership issue: How to frame AI trust as a question of accountability, not technology, and why speed without defensibility puts a CFO’s credibility at risk. - Context over cleverness: Why reliable AI depends on governed, proprietary business context and a clear line between deterministic calculations and probabilistic outputs, starting from the fragmented data and spreadsheets you have today. - Humans in the loop by design: Practical principles for deciding where AI acts, where it recommends, and where a person must review, so oversight scales with risk. - Sharper questions for any AI approach: The questions that separate dependable AI from risky AI, whether built in-house, added to a general-purpose tool, or bought, covering governance, auditability, security and proof of ROI.