EY UK: AI Project Sales Up 46%, General Consulting Down 3% - The Split Is Now Structural

The Two Businesses Inside One Firm

EY UK reported £3.86bn in fee income for the financial year ending 3 July 2026, a 2% rise from the prior year's £3.78bn. On the surface, respectable. One level down, the picture bifurcates sharply.

AI and data project sales rose 46% year on year. Consulting revenue fell 3%, extending a three-consecutive-year decline for that practice. EY-Parthenon, the firm's strategy and transactions arm, grew 6%; tax and assurance each added 3%. The firm is growing, just not via the service line that built its name in the 2010s.

CityAM reported that general consulting is in the red across the board: EY down 3%, KPMG down 3%, PwC's wider group down nearly 10%. Growth across the Big Four has come from mandatory regulatory work and AI-specific transformation engagements. Those two categories are structurally different businesses with different margin profiles, different resourcing models, and different competitive sets.

Partner Pay and the Productivity Paradox

Distributable profit before tax hit £702m, up from £679m. Average partner profit jumped 10.5% to £870,000, more than double the 5.2% average employee pay rise. EY hired 2,300 people during the year, made over 3,700 internal promotions, and raised the total value of staff bonuses by over 25%.

That staffing activity looks bullish. Read it alongside the consulting revenue decline, though, and a different story emerges: the firm is reweighting headcount toward AI-capable practitioners while volumes in traditional advisory compress. The 1.2 million learning hours logged by UK staff and the Microsoft Copilot rollout across the entire UK workforce are not culture initiatives. They are utilisation defences.

Globally, EY and Microsoft are committing £750m to build sector-specific AI tools, pairing Microsoft engineers with EY industry specialists. At that scale, EY is not buying a product. It is rebuilding delivery infrastructure.

Anthropic Sees the Same Gap, From the Other Side

The EY numbers land in the same week that the Financial Times reported Anthropic and London-based private equity firm Hg are creating an independent European AI consultancy, due to launch before the end of 2026, anchored by hubs in London and Berlin. The venture would embed AI engineers and consultants directly inside client companies. According to people familiar with the plans cited by the FT, the partners are targeting a business worth more than $10bn within five years.

This is a direct move into territory EY, Accenture, and the other large firms have been trying to own. Anthropic has not previously backed a standalone consultancy in Europe. Hg already employs around 150 AI specialists serving its portfolio and those remain in place, so the new venture is additive capacity, not a reorganisation of existing resource. The consultancy will work with Hg-owned businesses and external B2B software firms and other white-collar sectors.

The strategic logic for Anthropic is transparent: every implementation engagement converts model usage into recurring enterprise spend and locks in Claude as the operational layer. It is less a consultancy play than a distribution strategy wearing a consulting coat.

What the Trademark Data Signals

AI Business Dispatch analysis of IPO (TMD) data, as of October 2026, shows UK Nice Class 41 trademark filings, covering education, training, and entertainment services, hit just 2,909 in Q4 2026, down 67.8% on the prior period. Class 41 is where corporate training propositions, professional development platforms, and edtech brands register. A collapse of that magnitude in a single quarter either signals a genuine pull-back in new market entrants or a classification shift, as AI-native learning services migrate toward Class 42 (AI as a service, now formally recognised under the 13th edition of the Nice Classification, effective January 2026).

Either reading is uncomfortable for anyone building a standalone training brand. Fewer filings means less competition, yes, but also less investor confidence in the category as a standalone commercial proposition.

MBB Presses on Pay, Too

Management Consulted published salary data this week confirming McKinsey, BCG, and Bain have raised US starting salaries for the 2027 hire class, the first increase since August 2022. Post-MBA hires at all three firms now start at $200,000 base; pre-MBA hires at $120,000. BCG made the largest relative adjustment, having previously trailed McKinsey and Bain by $2,000 on both bands.

The driver, per Management Consulted, is renewed competition from big tech and AI labs, both of which are hiring again after two years of cuts. McKinsey is reportedly planning to take on 12% more new consultants in 2027 than this year. The talent war for analytically trained graduates is reigniting at precisely the moment EY's UK data shows consulting revenue still shrinking.

Something has to give. Either consulting demand rebounds faster than expected, pushed by AI implementation complexity, or the MBB salary raises squeeze margin in a market that is not yet delivering the revenue to absorb them. Partners at every tier will be watching utilisation data more carefully than press releases this quarter.