Finance teams are building their own AI tools without writing code

Moving from manual closes to live forecasting

OpenAI detailed how its internal finance team transitioned from manual spreadsheet processing to an AI-native infrastructure. Using tools like ChatGPT Enterprise and Codex, employees with no coding background are building interactive dashboards and custom tools. The ultimate goal is continuous forecasting and a zero-day close (giving leadership a real-time, reconciled view of the company's financial position instead of scrambling at the end of the month).

Finance professionals turn into infrastructure builders

The major shift is not just faster calculation, but a fundamental change in the nature of the work. About a fifth of the finance team's specialized AI use at OpenAI now resembles engineering. Instead of manually extracting and reconciling data across different systems, finance professionals build their own context-aware AI agents, such as IR-GPT for handling investor diligence. The team spends less time assembling inputs and more time analyzing scenarios and making decisions.

Strict ROI replaces the AI experimentation phase

The speed of deployment requires rigorous controls. OpenAI CFO Sarah Friar acknowledged that the era of unstructured AI experimentation is over. Tools must operate with clear access limits, budgets, and approval thresholds. While the model can instantly draft a report or flag exceptions, the final context, judgment, and authorization remain strictly human. The deciding metric is no longer merely deploying AI, but measuring the actual return on investment and cost for every completed task.

The true test lies in verifiable business value

The proof of a successful transformation won't be smarter spreadsheet macros, but how quickly executive leadership can react to market changes. Genuine adoption will depend on maintaining strict data governance, ensuring AI does not compromise accuracy, and demonstrably lowering operational costs compared to legacy workflows.

Lilith's verdict

The transition from static spreadsheets to live agents ends the CFO's role as a corporate historian. Accounting departments are quietly turning into internal development studios, leaving those who still manually mine data a month behind the market.