OpenAI’s CFO Says Finance Jobs Now Require AI Agent Skills Alongside Excel

Sarah Friar, OpenAI’s first chief financial officer, has a blunt message for finance professionals: knowing spreadsheets alone won’t cut it anymore. In a recent interview, she said she wouldn’t hire a finance person today who doesn’t know how to use a tool like Codex, the company’s AI coding agent. The comment, made at the Liquidity Summit, signals a major shift in what finance leaders expect from new hires—and how deeply OpenAI itself relies on its own technology.

Codex started as a coding tool but has grown into a platform used by more than five million people each week. About 20 percent of those users are knowledge workers—analysts, marketers, investors, bankers—and that group is growing more than three times as fast as developers. Inside OpenAI, non-technical teams use Codex to build internal apps, prepare executive materials, generate dashboards, and turn creative briefs into brand-ready output.

On Tuesday, OpenAI released a series of plugins that tailor Codex to specific jobs. A public equity investing plugin helps investors review earnings and test investment theses using data from Moody’s, FactSet, and others. An investment banking plugin assists with pitch materials and comparable analysis. More plugins for corporate finance, legal, and strategy consulting are coming soon. The company also plans to fold Codex into its flagship ChatGPT product.

The timing matters. Rival Anthropic has launched its own AI agents for banks and financial firms, handling tasks from drafting pitch decks to reviewing financial statements. Both companies are racing to win over enterprise clients as they prepare for potential public listings.

Friar has been building out her finance team to handle the demands of a company burning through enormous sums on computing infrastructure while readying for an IPO. She hired a new chief accounting officer and a corporate finance lead. The Wall Street Journal reported last month that she has begun informal talks with banks and moderated some of the company’s more ambitious spending projections.

Compute scarcity remains a binding constraint. OpenAI’s computing capacity has grown from 0.2 gigawatts in 2023 to roughly 1.9 gigawatts last year, and the company plans to lock in around 30 gigawatts over the next five to seven years—enough to power dozens of nuclear plants. Friar expects the shortage to persist through 2026, despite heavy upfront investment.

Finance leaders outside OpenAI are paying attention. Deloitte’s 2026 survey of more than 1,300 global finance executives ranked AI and automation skills as the top priority for talent development, outpacing traditional strengths like regulatory compliance. Friar’s own experience at OpenAI reflects the gap: when she arrived, some team members were experimenting enthusiastically with AI, while recent hires from traditional firms had barely touched it.

Still, the technology doesn’t replace judgment. Friar emphasizes trust, community relations, and sound capital allocation. Raised in Northern Ireland during the Troubles, she earned an engineering degree before working at McKinsey, Goldman Sachs, and leading finance at Square and Nextdoor. Those experiences taught her how to guide high-growth companies through complexity.

The bar has moved. Excel fluency once defined competence in finance. Now that standard includes the ability to direct AI agents that write code, pull data across systems, and recommend actions. Finance professionals who ignore the shift will find fewer opportunities at leading AI organizations. Those who embrace it may see their roles expand from scorekeepers to strategic partners shaping business outcomes at unprecedented speed.