Complete AI Training
Sign inGet my AI kit

Your job's AI kit

Get your AI kit

Tell us who you are and what you do. We show you your kit right away and email you the link: skills, prompts, AI agents, MCP servers and courses for your job.

500+ jobs ready, and we make a kit for any other job. No payment needed to look.

Share

AI news ·

PwC and OpenAI partner to deploy agentic AI for CFOs

PwC and OpenAI are deploying supervised finance AI after a $290,000 Deloitte penalty. OpenAI's agents processed five times more contracts without adding headcount.

1 min read
Finance

PwC and OpenAI are introducing agentic AI to finance departments, targeting planning, forecasting, and reporting workflows. The partnership moves the technology from a controversial experiment to an operational tool, addressing previous reliability concerns by keeping human supervisors in the loop.

Finance leaders remained cautious after a $290,000 judgment against Deloitte Australia in October 2025 over generative AI hallucinations in a government report. To mitigate those risks, the PwC and OpenAI deployment requires human employees to supervise the AI agents while the software executes and coordinates tasks. This structure aims to catch errors before they reach final reports.

Real-world deployment

OpenAI is acting as customer zero, using its own ChatGPT and Codex tools to run internal financial operations. According to the company, these agents "monitor payments, review contracts, update forecasts, and prepare reporting materials." PwC is now taking those internal lessons and applying them to external corporate clients.

Finance teams looking to understand these new workflows can explore an AI Learning Path for CFOs to see how the technology changes daily responsibilities. OpenAI officials said their agents processed five times more contracts without adding headcount and handled over 200 investor interactions during a single fundraising event.

Adding competing models

PwC also partners with OpenAI competitor Anthropic, giving clients access to the Claude AI assistant, a development covered in broader AI for Finance discussions. The firm sees strong interest from financial services, pharmaceutical, and life sciences companies. In the insurance sector, underwriting cycles could drop from weeks to days, while cybersecurity agents can respond to threats in minutes instead of hours.

Why this matters for finance professionals

The shift to agentic AI requires finance leaders to design workflows where machines execute tasks and humans verify outputs. CFOs must build supervision protocols that catch errors before they affect financial reporting or regulatory compliance.

Share