OpenAI's Finance Team Embraces AI

Building a finance function from the ground up at a hyper-growth company like OpenAI presented a unique opportunity to embed artificial intelligence at its core. The lessons learned from this process, detailed in a recent OpenAI News post, offer a roadmap for CFOs looking to transform their departments.

The ambition is to move beyond simply accelerating existing tasks, like closing books faster or refreshing forecasts more often. Instead, the goal is a real-time financial view that helps leaders act sooner and gives finance teams more capacity for strategic input. This involves redesigning work around consequential decisions, enabling experimentation, and establishing clear accountability.

From Manual to Real-Time

The journey began with familiar challenges: manual, recurring work for financial closing and forecasting. Even with access to cutting-edge AI tools, the team aimed for a "zero-day close", a real-time, reconciled, and traceable financial picture. This is complemented by continuously updated forecasting, showing business trajectories and the potential impact of decisions.

This shift means moving away from static spreadsheets and manual data searches toward live tools that integrate the business's full data context. It also empowers finance professionals to become builders of their own tools, extending their expertise.

Five Lessons for AI-Native Finance

The OpenAI finance team distilled their experience into five practical lessons:

1. Give everyone access, then create a reason to use it: Broad access to AI tools, paired with structured experimentation on real problems, drives adoption. A finance hackathon, for instance, led to custom GPTs like IR-GPT for investor relations, turning abstract capabilities into practical solutions. This requires both bottom-up ideas and top-down strategic focus.
2. Redesign the full workflow around the decision: Instead of just assembling inputs for decisions, finance leaders can redesign the entire path from raw data to final choice. For the close, this means connecting spending plans, actuals, and transaction details into a continuously reconciled view, with AI preparing initial explanations and flagging exceptions. This foundation supports interactive forecasting and scenario planning, allowing leaders to see what changed, why, and how decisions could alter outcomes. The same principle applies to capital allocation, enabling dynamic adjustments based on return data.
3. Finance professionals become builders: The most significant transformation is enabling finance professionals to build the tools they need. Using tools like ChatGPT and Codex, team members are creating live dashboards and custom AI applications. This moves work from static models to dynamic tools that respond to questions and update with changing data. A teammate, for example, built a tool to convert monthly ad forecasts into weekly and daily plans, accounting for various factors.
4. Pair speed with clear accountability and controls: AI accelerates tasks like drafting investor diligence responses, but human judgment remains central. Clear governance is essential: defining data access for AI, setting action limits, and establishing escalation paths. Every AI output must connect to a reliable source, and changes to approved baselines require finance authorization. Implementing usage limits, budget controls, and role-based access allows for disciplined AI management.
5. Measure value per unit of intelligence: Success is measured not by AI usage but by operational impact. For each workflow, questions focus on whether AI completed important work, its total cost (including human time), the quality of the result, and whether it improved speed or decision-making. For the close, metrics might include cycle time and automatic reconciliation rates. For forecasting, it could be accuracy and the time to produce new scenarios.

The CFO's Strategic Mandate

Finance sits at the nexus of strategy, capital, data, and performance. This positions CFOs uniquely to lead an AI transformation. An AI-native finance function promises faster cycles, stronger controls, better decisions, and more time for critical judgment. The path forward involves identifying a consequential workflow, providing the tools, supporting redesign, maintaining accountability, and diligently measuring outcomes. The ultimate ambition is a finance team that understands real-time operations, anticipates future possibilities, and helps the company make better decisions sooner, turning intelligence into durable value.