OpenAI Automates Finance With AI Across Treasury, Tax, Reporting

PwC and OpenAI are collaborating to reshape financial operations, beginning with a pilot program within OpenAI’s finance organization. The partnership focuses on practical application, building AI agents to automate workflows across functions including treasury, tax, and investor relations; this approach positions OpenAI as its initial client for real-world validation. The companies are currently developing a procurement agent within OpenAI’s finance department, with learnings intended for rapid deployment to other clients. “Finance is at an inflection point, where organizations are moving from process efficiency to intelligent, decision-centric operations,” says Tyson Cornell, US Advisory Leader, PwC; through this collaboration, they aim to help clients embed agentic AI into their finance functions, enabling proactive insights and stronger controls.

PwC and OpenAI Deploy “Customer Zero” Finance Agents

OpenAI is applying its technology to its own finances; the company’s finance division is piloting artificial intelligence agents across functions including investor relations, treasury operations, tax compliance, financial reporting, corporate development, and contract review, providing valuable, real-world implementation data. Unlike conventional software launches, PwC and OpenAI are prioritizing a “build in the real world” approach, starting with these agents within OpenAI’s financial infrastructure before offering them to external clients. This allows for rapid iteration and refinement based on immediate operational feedback, accelerating the development of practical applications. The collaboration extends beyond conversational AI; OpenAI’s Codex, alongside ChatGPT, is being leveraged to construct and customize essential financial tools, including dynamic dashboards and detailed spend trackers.

These tools are designed to automate repeatable tasks, integrate data from disparate systems, and improve decision-making with robust governance and human oversight. Agents are being utilized to monitor payments, scrutinize contracts for policy adherence, dynamically update financial forecasts, and proactively identify risks before critical reporting periods. OpenAI’s finance team has quantified benefits from this early adoption, reportedly processing five times more contracts with the same staffing levels using Codex, and managing over 200 investor interactions during a recent fundraising round with the assistance of a tool called IR-GPT. PwC contributes its expertise in financial transformation, internal controls, and implementation to translate these prototypes into scalable, production-ready solutions for enterprise environments. As agentic workflows expand, both companies recognize the need for comprehensive visibility into AI usage, token consumption, and associated costs, ensuring responsible and manageable adoption across organizations.

Codex and Workspace Agents Automate Core Finance Workflows

The convergence of artificial intelligence and financial operations is now focused on automating core workflows with agents built on models like Codex and ChatGPT. While many firms explore AI’s potential, OpenAI is uniquely positioned as its own primary test case, deploying these agents internally across functions including investor relations, treasury, tax, reporting, corporate development, and contract review, effectively functioning as an initial client. This internal implementation provides invaluable operational experience before any external product launch, allowing for rapid refinement based on real-world demands. PwC and OpenAI are taking a practical approach, currently piloting a procurement agent within OpenAI’s finance organization before wider client rollout. This “build in the real world” strategy prioritizes demonstrable results over theoretical design, a departure from traditional software development cycles.

The collaboration extends beyond procurement, encompassing agents designed to handle planning, forecasting, reporting, payments, and the complexities of the accounting close. These agents aren’t intended to replace human oversight, but rather to augment it by monitoring payments, reviewing contracts for policy compliance, and proactively surfacing risks before critical deadlines. Workspace Agents then facilitate repeatable workflows across familiar platforms, while Skills and Connectors ensure adherence to established processes and access to relevant enterprise data, aiming to transform how CFOs operate.

Finance is at an inflection point, where organizations are moving from process efficiency to intelligent, decision-centric operations. Through our collaboration with OpenAI, we’re helping clients embed agentic AI into the core fabric of the finance function, enabling more proactive insights, stronger controls, and a more adaptive operating model.

Tyson Cornell, US Advisory Leader, PwC
Ivy Delaney

Ivy Delaney

We've seen the rise of AI over the last few short years with the rise of the LLM and companies such as Open AI with its ChatGPT service. Ivy has been working with Neural Networks, Machine Learning and AI since the mid nineties and talk about the latest exciting developments in the field.

Latest Posts by Ivy Delaney: