The billable hour is dying
Janet Truncale just confirmed the autopsy. Speaking at Goldman Sachs today, the EY Global Chair and CEO outlined a pivot that should terrify every junior associate in London and New York. The firm is not just using AI to write emails. It is scaling agentic AI. This is a fundamental shift in the architecture of professional services. It moves beyond passive large language models into the realm of autonomous execution.
Labor is the largest expense. Code is a fixed cost. For a firm like EY, the math is simple. If an agentic system can handle the grunt work of a mid-market audit, the margin expansion is astronomical. Truncale’s appearance at the Talks at GS series signals to the market that the Big Four are no longer service providers. They are becoming software-defined enterprises. This is the industrialization of white-collar work.
The mechanics of the agentic shift
Agentic AI differs from standard generative AI through its capacity for tool-use and multi-step reasoning. Traditional LLMs are reactive. You prompt, they respond. Agentic systems are proactive. They are given an objective, such as reconciling a complex multi-currency ledger, and they determine which tools to use to achieve it. They can access APIs, query databases, and cross-reference regulatory frameworks without human intervention. This is the ReAct (Reason and Act) framework applied to global finance.
According to recent reports from Reuters, the deployment of these agents at EY is already impacting how client services are billed. We are seeing a move away from time-based metrics toward outcome-based pricing. The efficiency gains are being captured by the firm, not passed to the client. This is the cynical reality of the automation of expertise. The firm reduces its headcount while maintaining its fee structure.
Professional Services Automation Trends
Projected Task Allocation in Global Professional Services
The Goldman Sachs endorsement
Goldman Sachs does not host these talks for the sake of academic curiosity. They are looking at the efficiency ratio. When Truncale discusses scaling AI, she is speaking to the institutional investors who trade Goldman’s stock. They want to see how professional services firms will survive a world where the cost of intelligence is trending toward zero. The endorsement from the bank suggests that the capital markets are now pricing in this transition.
The risk is concentration. If EY successfully scales agentic AI, the barrier to entry for smaller firms becomes insurmountable. The data moats are too deep. The compute costs are too high. We are entering an era of the ‘Super-Firm’ where a handful of entities control the automated infrastructure of global commerce. As noted in a Bloomberg analysis yesterday, the market capitalization of firms leading in agentic deployment has outpaced the broader S&P 500 by 14 percent since January.
The human in the loop fallacy
Management often uses the phrase ‘human in the loop’ to soothe labor concerns. It is a marketing term. In practice, the human becomes a rubber stamp. When an AI agent processes ten thousand transactions in seconds, a human reviewer cannot possibly provide meaningful oversight. They are there for legal liability, not for quality control. The technical reality is that the agent is the primary actor. The human is the legacy interface.
This creates a training vacuum. If the agents are doing the work of junior associates, where do the future partners come from? You cannot manage what you do not understand. By hollowing out the bottom of the pyramid, EY and its peers are taking a massive gamble on the long-term viability of their leadership pipeline. They are trading future talent for immediate margin.
Capital allocation and the compute war
Scaling agentic AI requires more than just clever algorithms. It requires massive infrastructure. EY’s partnership with major cloud providers is no longer a utility arrangement. It is a strategic alliance. The firm is essentially renting the brains of the global economy. This shift in capital allocation from payroll to compute is the defining trend of the 2026 fiscal year.
Investors should look closely at the upcoming SEC filings for the major consulting groups. The line items for ‘Technology and Infrastructure’ are expected to eclipse ‘Personnel Expenses’ for the first time in several sub-sectors by the end of the month. This is the moment the service industry becomes the tech industry.
The next data point to watch is the June 30 quarterly performance review. If EY reports a significant drop in headcount alongside a rise in revenue per employee, the transition is complete. The agentic era is not coming. It is already here. Watch the July 15 earnings calls for the first concrete ‘Agentic Efficiency’ metrics from the broader sector.
