OpenAI CFO Sets Out Five Lessons for AI-Native Finance

When Sarah Friar joined OpenAI two years ago, the company had a small finance team supporting rapid growth and still relied on recurring manual work to close its books and refresh forecasts. Friar says an AI-native function matters because finance can shift from reconstructing past performance to seeing the business in real time and shaping decisions earlier. OpenAI set two ambitions: a “zero-day close” and automated, continuously updated forecasting, though both remain works in progress.

On Aug. 10, 2026, OpenAI published Friar’s five lessons: give employees broad access and structured reasons to experiment; redesign end-to-end workflows around decisions; turn finance professionals into builders; pair speed with accountability and controls; and measure value per unit of intelligence. OpenAI research found 40% of finance professionals’ specialized AI use falls outside traditional finance and 22% involves engineering tasks. Friar said a controlled IR-GPT cut first-draft responses from hours, sometimes overnight, to seconds. No investment amount was disclosed.