PwC vs. EY: The Big Four Make Opposing Bets on AI
Anthropic just published an AI Exposure Checklist that maps every major profession against two numbers: the theoretical percentage of the job AI could automate and the observed percentage it’s automating right now.
Business and finance (i.e., accounting) sit at the outer edge of the chart.
David Leary and I talked about it on episode 479 of The Accounting Podcast, and it sparked an interesting conversation because the same week we dug into that chart, two of the biggest firms in the world made opposite bets on what it means.
The Big Four are Split on the Future of Your Job
PwC’s U.S. CEO told The Guardian that partners and staff who haven’t fully adopted AI risk will be pushed out of the firm. They cut 5,600 employees last year and shifted hiring towards data specialists and engineers.
They’re building AI-powered tools and charging clients a subscription fee to access them, no PwC professional required. That’s what they’re calling the future.
EY, meanwhile, just doubled its CPA exam pass bonus to $10,000.
One firm is betting on machines. The other is betting on people.
As we discussed on the pod, the problem with PwC’s model is that when you sell an AI tool with your logo on it, you’re competing against every other AI tool, including the one the client already has on their laptop.
The value of the Big Four brand was never “we have software.” It was “we stand behind the answer.”
Is human judgment really that easily replaceable?
Automating it away doesn’t scale the business model. It undermines the reason clients pay a premium.
EY doubling down on the CPA is a bet that credentialed humans (those who can be held accountable and understand the work the AI is doing) will be worth more, not less.
The Subtext in the Anthropic Chart
The Anthropic chart actually tells a more nuanced story than the headlines suggest. Yes, business and finance has a near-ceiling theoretical exposure. But the observed exposure (what’s being automated today) is still low.
David has been asking on LinkedIn for weeks whether anyone can name one accountant they personally know who lost their job because a company implemented AI. Still at zero.
History shows that labor-saving technology doesn’t tend to reduce work. It raises standards and creates more work. Your clients won’t need fewer accountants. They’ll expect more from the ones they have.
Know Your Level, Then Skip One
I left public accounting as a manager. Give that version of me today’s tools, and I’m doing partner-level work inside of six months.
That’s not a brag. It’s what AI does to the experience curve. It compresses a decade of knowledge accumulation into a much shorter climb.
AI is going to blow up the staff-to-partner ratio. A big firm running a large client engagement today might assign 10 people to it. In a few years, that’s a team of three.
In small firms, one person will routinely do what used to take a team.
That changes how you should be thinking about your career right now.
If you’re a staff accountant, stop thinking like staff. Use AI to produce manager-level output: analysis, judgment calls, and client communication. Do it before the headcount reductions catch up with you.
If you’re a manager, the same logic applies. Push yourself to director or partner-level work.
If you’re a partner, you have two options. Use AI to develop your staff faster, or use it to run leaner.
Both work. Doing neither doesn’t.
