Good morning. I sat in on a webinar yesterday where an OpenAI exec casually mentioned that his entire department works with an AI agent—and the more I listened, the clearer it became that the tool was almost beside the point. The real story is what OpenAI’s finance team is being asked to become.
The finance department increasingly has the mandate to help leadership act while the outcome can still change. Within OpenAI’s finance department, led by CFO Sarah Friar, there is a strategic finance team that functions as a cross-functional partner to product, partnerships, engineering, and go-to-market teams. The team’s tool of choice is ChatGPT Work, an agent that can act across apps and files, stay on a task for hours, and turn a goal into a finished product.
The webinar was led by three members of OpenAI’s strategic finance team: Stephanie Struck, head of product finance; Kyle Kober, director of product finance; and Jackson Wang, who leads data science. “Our entire finance department uses ChatGPT Work,” Kober said. Teams combine approved financial data, Excel models, and business context to get frontier-level intelligence that completes tasks two to three times faster, he said.
Struck demonstrated the tool auditing quarter-end materials to catch discrepancies between board decks and actuals. Wang described shifting forecasting “from more of a spreadsheet-based system to more of a statistical-based forecasting model,” improving both accuracy and collaboration. They detailed how the tool automates monthly close by pulling data from systems like Databricks into one validated dashboard. ChatGPT Work speeds up and structures the close, but doesn’t replace the underlying accounting system or a controller’s sign-off.
Newer features include dynamic slide editing, a finance assistant for quick questions, an on-the-go podcast, with a “CFO spotlight” that surfaces the month’s top priorities and biggest variances—a feature the presenters said Friar values.
Regarding data governance, Kober said: “All of it is governed and auditable with controls to lock down the spend and the usage of your ChatGPT work experience.” However, there wasn’t much discussion of who owns the data, or how agent decisions get audited.
The skills that matter now
Struck addressed what OpenAI looks for in finance hires. Accounting, forecasting, controls, and analytical rigor are still required, she said, but they’re now baseline, not differentiators. What sets people apart is comfort with data and AI tools paired with judgment: knowing which questions to ask, challenging an AI’s output, and translating it into a business implication. “Why does this matter for the business?” is the question she wants her team asking constantly. Equally critical is communication—insights only matter, Struck said, if they’re shared quickly enough to change how the organization actually operates.
It seems that OpenAI isn’t betting that the tool alone gets this right. It’s betting on people who know when to push back on what it says.
Sheryl Estrada
Sheryl.Estrada@fortune.com
This story was originally featured on Fortune.com