OpenAI CFO Sarah Friar said AI is already reducing the need for some workers inside the company, especially on repetitive tasks.
In a Tuesday interview with CNBC's Jim Cramer on 'Mad Money,' Friar used OpenAI's finance team as an example. She said the company is doing things with its own products that 'do limit the number of people we need,' but added that in many cases those roles are not 'great jobs.'
Friar pointed to credit checks as one example. OpenAI's procurement operation was conducting them at an annual rate of 2,800 last year, she said. 'That credit check is a pretty mundane piece of work,' Friar said. 'Do you want a junior analyst doing that, or can we get a very intelligent model to do it for us?'
She said the goal is to let people do more work 'at the edge,' where intelligence is needed, while the model handles 'the rote stuff' at better scale and with more accuracy. Friar said the financial payoff was significant: the cost per check fell from $200 to about 17 cents.
Her comments come as some leading AI figures have warned much more sharply about employment. Anthropic CEO Dario Amodei has said AI could eliminate half of entry-level white-collar jobs. Geoffrey Hinton, the Nobel Prize-winning computer scientist known as a 'godfather of AI,' recently said such predictions were 'probably right.'
Hinton also said AI still fails at some tasks, but warned that the technology is improving quickly. 'We're in the very early stages,' Hinton told Bernie Sanders in an interview in November 2025. 'This is a new technology. It's getting better very fast.'
For founders, the message is simple: AI is first coming for repetitive back-office work, not the highest-value judgment calls. That can mean lower costs, faster operations, and cleaner margins if you use the tools well.
But the other side of the playbook is just as clear. When a company can replace a $200 task with a 17-cent one, the pressure on headcount will spread quickly beyond finance. The winners will be the teams that shift people toward work customers actually pay for, instead of paying salaries to automate the mundane.
