The Big Four firm says AI solves a talent shortage. PwC just cut graduate hiring by a third partly because of AI. Those two facts do not fit together comfortably, and nobody in the industry has answered the obvious follow-up question.
image from grok
EY is deploying its Nvidia-built EY.ai Agentic Platform to global audit teams, supporting 80,000 employees with 200 specialized agents and achieving 50% efficiency gains in audit processes, including expanding transaction sampling from 100 to 10,000 transactions. Stanford research shows entry-level AI-impacted job postings fell 16% in two years, while PwC has explicitly linked a planned one-third reduction in US graduate hiring to AI's impact on associate-level work, creating a paradox with EY's stated commitment to maintain human workforce levels while training AI for new hires.
EY is rolling out advanced AI agents to its global audit teams this week, a deployment the firm says will reshape how financial statement audits are conducted over the next two years. But the humans at the bottom of the professional services pyramid are feeling the squeeze first.
The Big Four firm is deploying its EY.ai Agentic Platform, built with Nvidia, to audit teams worldwide the week of April 6, 2026, according to Bloomberg Tax. The platform already supports 80,000 EY employees with roughly 200 specialized AI agents. The firm has advanced 1,000 AI agents into development or production in 2025 and aims to scale to 200,000 agents by 2028, Business Insider reported.
The efficiency gains are concrete. An audit process that once took five hours can now be completed in two and a half hours with AI support, EY told Business Insider. Transaction sampling, a core audit procedure, has expanded from 100 transactions to 10,000. Audit is the kind of work where efficiency gains are visible and measurable — which is why it is the proving ground for what AI can do to a professional services business.
What AI can do to entry-level accountants is less clear, and less comfortable.
The paradox EY cannot quite resolve is this: the firm says it is solving a talent shortage by training AI into new hires, while simultaneously reducing the work those new hires are designed to do. Raj Sharma, EY's global managing partner for growth and innovation, told Business Insider the firm has no plans to reduce its human workforce. "Under no circumstances are we ever thinking that we will get rid of our human workforce and do this just through AI agents," he said. The firm generated just over $50 billion in revenue last year, employs roughly 400,000 people, and plans to hire 1,600 junior staff this year. All Big Four firms are planning to hire at least 1,000 juniors each.
But PwC tells a different part of the story. The firm has reduced its US graduate hiring goals, in part because of AI's impact. An internal presentation obtained by Business Insider showed PwC US planned to cut graduate hiring by a third over the next three years, with a bullet point explicitly linking the decision to AI's effect on associate-level work. Errol Gardner, EY's global head of consulting, said EY has made no material change to graduate recruitment — but that answer sidesteps the question of whether the quality and nature of entry-level work remains what it was.
The data is catching up to the corporate messaging. A Stanford study published in 2025 found that hiring for entry-level, AI-impacted jobs — including junior accounting roles — fell by 16 percent in roughly two years. A separate Stanford finding, cited by Fortune, showed that workers in the most AI-exposed occupations experienced a 6 percent employment decline from late 2022 to July 2025, while workers aged 35 and older saw 6 to 9 percent employment growth over the same period. The youngest workers in the most automatable roles are losing ground. The oldest workers are gaining it.
Jess Larsen, founder of the workforce development firm Thriving Humans, frames the problem as structural rather than cyclical. "Managers have always found it hard finding effective work for apprentices and entry-level positions to do," she told ICAEW. "AI is removing that work and making a problem that was already there worse." The traditional consulting and accounting pyramid — an engagement manager plus three analysts and associates — was designed in part to give junior staff things to do while senior staff did the thinking. If AI is now doing the junior work, the pyramid's purpose dissolves.
Yvonne Hinson, CEO of the American Accounting Association, has been asking the question nobody in the industry has answered: how do you develop foundational professional skills without grunt work when AI handles the grunt work? "This is the big question right now that I haven't been able to get anybody to answer for me," she told Business Insider.
EY's public position is that AI handles the repetitive work and lets humans focus on judgment calls that require context, relationships, and accountability. Dan Diasio, EY's global consulting AI leader, and Errol Gardner argued in a Business Insider interview that entry-level staff arrive without accumulated assumptions about how work should be done — a blank slate that is, paradoxically, an advantage in a world where AI can surface patterns human experience would miss. The framing is that junior staff are not being displaced; they are being elevated.
The counterargument is not abstract. The professional services industry has always used entry-level roles as a training ground: the grunt work was also the education. Audit sampling 100 transactions teaches pattern recognition. Reviewing flagged discrepancies builds judgment. Sitting in client meetings and watching a partner navigate a difficult conversation is how junior staff become senior staff. If AI handles all of that, the question is not whether the firm needs fewer juniors this year. The question is what the partner population looks like in fifteen years, and whether anyone who got their formative experience from an AI agent can validate what the AI agent did.
EY is investing more than $1 billion annually in AI platforms, and AI-related revenue was up 30 percent in fiscal year 2025. The firm has added 61,000 technologists since 2023; they now make up roughly 15 percent of its total workforce. These are real numbers attached to a real transformation. The agents exist and are running in production. The audit teams are being touched.
What is less clear is whether EY — or any Big Four firm — has actually answered the Hinson question. The talent shortage and the AI deployment are happening simultaneously. The graduate hiring continues. The grunt work shrinks. The pipeline for developing senior judgment is not yet broken, but the direction of travel is not ambiguous either.
The next two years of the EY.ai deployment, as the firm integrates autonomous agents into each step of financial statement audits globally, will be a test case that every professional services firm is watching. If the junior headcount holds and the work evolves, the industry's reassurance about AI being an elevation story will look prescient. If the grunt work disappears and the graduate numbers follow the PwC trajectory, the firms will have automated the job that used to train the people who now run them.
Story entered the newsroom
Assigned to reporter
Research completed — 13 sources registered. Wire signal traced to Business Insider GNews link (unverifiable paywall). EY AI agents at scale: 200 agents supporting 80,000 employees via EY.ai Agen
Draft (1109 words)
Approved for publication
Published (1068 words)
Get the best frontier systems analysis delivered weekly. No spam, no fluff.
Agentics · 1h 5m ago · 4 min read
Agentics · 1h 15m ago · 4 min read
