What building an AI-native finance function taught me
A finance leader at OpenAI describes how the company is redesigning its finance function around AI, aiming for real-time financial visibility and automated forecasting to enable faster, more informed decisions.
OpenAI’s finance team set two goals: a zero-day close to provide real-time, reconciled financial data, and continuous forecasting to show how the business is evolving. These ambitions aim to move beyond traditional month-end closes and static spreadsheets, replacing manual data gathering with live tools built on comprehensive business context. The approach seeks to give leaders immediate insight into financial position and potential future outcomes, enabling quicker, more informed decisions.
To achieve these goals, OpenAI emphasized broad access to AI tools paired with structured experimentation. A finance hackathon engaged teams in identifying tasks to automate, resulting in custom GPTs like IR-GPT for investor relations and tools for procurement and tax. The initiative demonstrated how frontline employees could rapidly prototype solutions, with technical experts accelerating development. The lesson for CFOs is to combine secure AI access with strategic prioritization, fostering bottom-up innovation aligned with top-level business priorities.
Finance teams traditionally spend significant time assembling data for decisions, from reconciling spreadsheets to preparing slides. AI shifts this dynamic by redesigning workflows from source data to decision-making. OpenAI’s zero-day close model connects spending plans, ledger actuals, purchase orders, and transaction details into a continuously reconciled view, with AI flagging variances for review. This eliminates the post-period scramble to reconstruct the business, replacing it with a live foundation for forecasting and scenario analysis.
The broader strategy involves starting with consequential decisions and mapping the data, tools, and approvals required to support them. OpenAI’s continuous forecasting integrates statistical models, sales data, operating metrics, and finance judgment into interactive tools. Leaders can inspect assumptions, compare scenarios, and assess the impact of decisions like capital allocation or marketing spend. The goal is automated forecasting that delivers faster scenario analysis and clearer ownership, improving the speed and quality of every decision cycle.