According to Fortune, during Thursday’s Emerging CFO virtual event with Workday, finance leaders revealed AI is fundamentally reshaping their departments but only when backed by clear strategy. James Glover from Deloitte warned companies deploying AI without broader plans are struggling to capture enterprise value, while implementations for agentic AI platforms take six to 12 months. Webflow CFO Craig Mestel shared his team now uses LLMs to automate routine policy questions, freeing analysts for higher-value work, while INRIX CFO Thadd Stricker reported AI has improved forecasting accuracy to 95% with less manual input. Greenlight’s Michelle Cheung said AI strengthens risk management and streamlines operations, and early adopters are starting to see “green shoots” of ROI despite many initial failures.
Strategy first, reality check
Here’s the thing everyone’s finally admitting: AI without strategy is just expensive tech theater. Glover’s point about defining objectives first—efficiency, control, effectiveness, or engagement—is something companies should have been doing from day one. But let’s be honest, how many finance departments jumped on the AI bandwagon because everyone else was? The fact that implementations take six to 12 months shows this isn’t some magic button you press. You actually have to train people, build comfort with simple tools first, and accept that this is a marathon, not a sprint.
Real-world stumbles
The most refreshing part of these CFO confessions? They’re openly admitting their failures. Stricker’s ChatGPT legal research producing confident but wrong case citations is exactly the kind of nightmare scenario that keeps compliance officers awake. And Mestel’s “it’s not great at math” admission about using ChatGPT as a junior analyst? That’s the reality check everyone needs. These aren’t theoretical problems—they’re happening right now in billion-dollar companies. The variance analysis struggles show that even seemingly straightforward financial tasks require serious iteration and testing before AI can handle them reliably.
Human oversight still critical
What’s fascinating is how these finance leaders are building guardrails while pushing innovation. Webflow partnering with HR to create AI performance review frameworks? That’s actually smart governance. The hackathons at Greenlight for better AI understanding? That’s how you build organizational literacy. But let’s not ignore the AuditBoard report findings they mentioned—only 28% of leaders are confident in auditing AI risks, and 63% haven’t defined formal risk frameworks. That’s a massive governance gap that could come back to haunt companies when regulators eventually catch up.
Broader CFO moves
Beyond the AI discussion, there were some significant CFO appointments worth noting. Steve Fieler moving from Google to lead Waymo’s finances brings nearly 30 years of experience to the autonomous vehicle space. Michael Nofi taking over at Edgewise Therapeutics brings biopharma accounting expertise, while James G. Mackey’s promotion at BankUnited and Martino Cadoni joining DeepL from Klarna show continued musical chairs in the finance executive world. Disney extending Hugh Johnston’s tenure through 2029 with a $16.5 million equity award? That’s some serious vote of confidence in turbulent times.
The big picture
So what’s the real takeaway? AI is delivering value in finance, but it’s messy, requires human oversight, and demands strategic thinking from the start. The “don’t give up” mentality Mestel advocates is crucial because the first attempt will probably fail. But here’s what worries me: are companies building the right risk frameworks while they experiment? The AuditBoard numbers suggest most aren’t. And when you’re dealing with financial forecasting, legal research, or transaction analysis, getting it wrong isn’t just inconvenient—it could be catastrophic. The CFOs who succeed will be those who balance innovation with iron-clad governance.
