KPMG Sees Revenue Drop for the First Time in Years as AI Reshapes the Back Office of the Big Four

KPMG UK reported on August 24 a 1% decline in annual revenue, down to £2.5 billion (approximately $1.79 billion) for its fiscal year 2026. This is the first significant contraction for the British firm in several fiscal years and coincides with a quarter when PwC, EY, and Deloitte announced workforce reductions. The official justifications differ in footnotes, but a common vector emerges across all transcripts: AI agents are accomplishing in seconds what a senior associate would take three days to complete.

KPMG had previously cut about 10% of its corporate services division in the UK, which equates to approximately 200 positions, in the first half of the year. PwC initiated a voluntary resignation program focused on senior associates and managers in UK auditing, though it has not publicly disclosed the number of departures. Combined with the nearly 400 exits driven by KPMG following the ban on UK government contracts and the prior restructuring wave at Deloitte, the sector has entered net contraction in headcount for the first time since the post-pandemic boom.

This article continues The New Times series on how AI is reshaping careers across the global professional services landscape. Following an analysis of strategy consulting and banks, the next exposed level is auditing, with KPMG's data marking a critical downturn.

Why Auditing Fell First

The auditing sector is focused on tasks that AI agents currently perform best. Extracting structured data from invoices, conducting sampling tests, reconciling balances, and verifying policies against exception tables are tokenizable tasks with definitive right or wrong answers, backed by extensive historical data for fine-tuning. Deloitte, EY, PwC, and KPMG are scaling AI agents in audit, tax, and advisory functions in parallel, all converging on a twelve-month timeline for automating substantial portions of substantive testing.

The impact on the pyramid is severe by design. The auditing contract historically compensated senior partners through margins generated by a workforce of associates billing at hourly rates. If associates become operators of AI agents, billable hours decrease, engagement pricing drops, and firms must resort to layoffs to maintain partner margins. This is the same mechanism previously described for strategy consulting and banking, now with an additional variable: the auditing regulator is monitoring in real time.

The Counterargument Has a Name

The opposing argument warrants honest attention, and the most rigorous analysis has emerged from within the legal sector. When Baker McKenzie announced cuts of between six hundred and one thousand back-office positions in February citing AI, a wave of legal commentators argued that the firm was using AI as a scapegoat for restructuring driven by declining profitability and over-hiring post-pandemic. An article from Above the Law in February elaborated on this argument, while Law360 Pulse revisited the issue in July, asserting that it is still premature to attribute the cycle of layoffs in the professional services sector predominantly to AI.

Evidence weakening the central thesis of this article is evident. Auditing and consulting firms overhired relative to demand between 2021 and 2023 and had been operating with an unusually large bank of hours. Part of the cuts in 2026 addresses this excess even in a scenario without any AI agents running. The distinction that shallow debates often confuse is that which separates a cyclical correction of headcount from a structural erosion of the pyramidal model. Both phenomena are occurring simultaneously, with the latter being the one that will fundamentally alter the business.

The Point That Only Appears in the Organizational Chart

The point that the official communication from the Big Four has not yet voiced is the issue of the talent funnel. If substantive testing becomes automated, first-year associates will no longer perform the tasks that train the senior managers of eight years down the line. This is the same warning given by Debasish Patnaik, senior partner and leader of QuantumBlack AI at McKinsey, regarding banks, stating that classes of junior analysts are being reduced by up to two-thirds. Auditing carries this dynamic with an additional complication: the professional standard requires a minimum number of supervised hours for auditor accreditation.

Where the Reading Matters Outside of London

In the UK, the drop in KPMG UK serves as both an anchor and a signal, while the Financial Reporting Council closely monitors the effect of automation on audit quality in listed firms. In the United States, PCAOB is engaged in a parallel discussion, and the American affiliates of the Big Four are currently testing the same stack of agents under less regulatory pressure. In India, home to the Global Capability Centers for all four firms, the effect is twofold. Part of the work that returned from offshore to automation is eliminating positions in Bangalore and Gurgaon; another part reinforces platform engineering required by the AI agents, raising the salary floor for technical profiles.

In Brazil, where the Big Four operate with a similar pyramid structure and concentrate work in São Paulo, Rio, and Belo Horizonte, pressure is arriving from two channels. The fee schedules of Brazilian big audits are already under review by multinational clients who replicate agreed prices from abroad. Moreover, the pipeline for trainees, historically one of the few structured entry points into the corporate market, is beginning to shrink.