OpenAI and PwC collaborate to reimagine the office of the CFO

Finance teams sit at the center of how organizations plan, allocate capital, manage risk, and make decisions. To help them keep up with growing demands, PwC and OpenAI are collaborating to help enterprises reimagine the office of the CFO with AI agents that can automate workflows, coordinate across systems, surface risks and insights, and support better decisions with strong governance and human oversight.

Together, PwC and OpenAI are building AI agents around the core operating rhythms of finance, from planning, forecasting, and reporting to procurement, payments, treasury, tax, and the accounting close.

What sets this collaboration apart is its focus on building in the real world, not just designing in theory. For example, PwC and OpenAI are building a procurement agent inside the OpenAI finance organization, and are applying those learnings to additional agents across core finance workflows.

Building agents around real finance workflows

OpenAI has already been applying AI agents across its own finance organization, using tools like ChatGPT and Codex to support workflows across investor relations, treasury, tax, reporting, corporate development, and contract review. Now, PwC and OpenAI are building on those learnings to help other CFOs apply AI agents across their organizations.

In practice, that means helping finance teams automate repeatable work, connect context across systems, and support better decisions with clear governance and human oversight. Agents can help monitor payments and exceptions, review contracts or invoices against policy, update forecasts as business conditions change, prepare reporting materials, and surface risks before month-end or quarter-end close.

OpenAI as customer zero

The OpenAI finance organization is serving as “customer zero”  for applying AI agents to finance work, using that production environment to validate enterprise-scale workflows, governance models, runtime controls, and human-agent collaboration patterns.