The consulting pyramid was never just an org chart. It was a business model: a wide base of junior consultants doing research, synthesis and slide production, billed by the hour, supervised by a narrowing tier of managers and partners. That base is contracting, and AI is one of the reasons. The 2026 edition of Spaik's State of AI in Consulting finds that AI is hitting hardest exactly where the pyramid is widest, and that three competing shapes are emerging to replace it.
Key takeaways
- The junior tier is most exposed because its core tasks, research, data synthesis and slide-building, are what AI does fastest and most reliably. Entry-level salaries are being held flat while revenue per employee rises.
- PwC cut headcount by about 1.5% from 2024 to 2025, dropped its target of adding 100,000 employees by mid-2026, and trained over 315,000 staff in AI. Its UK Chief People Officer acknowledged that a graduate trainee now does what a senior associate did three to five years ago.
- AI-native firms such as Queen's Tower Advisory and Unity Advisory run roughly 80% AI agents to 20% human consultants; Xavier AI builds AI consultants outright.
- Three structures are emerging to replace the pyramid: the diamond (wide AI-fluent middle), the obelisk (taller, leaner, three redefined roles) and the hourglass (senior advisors above an AI platform, governance below). All three thin the junior base and reward technical and governance skills.
Why the junior tier is the most exposed
Ask what a first-year analyst actually does in a week and the answer is a list of tasks that generative AI now performs in minutes: pull together desk research, synthesise expert-call notes, structure a first draft of the analysis, build and format the slides. None of those tasks disappears; each becomes faster and cheaper to produce. The consequence is not that firms fire analysts. It is that they need fewer of them per engagement, and the ones they keep are asked to do something different.
The report identifies a second, quieter signal: firms are keeping entry-level salaries flat as revenue per employee rises. The productivity gain is real, and it is not being shared with the tier that used to generate the leverage.
What PwC's numbers show
PwC cut headcount by about 1.5% from 2024 to 2025 and dropped its goal of adding 100,000 employees by mid-2026, while training over 315,000 staff in AI. Global Chair Mohamed Kande has said the training is “boosting the productivity of our people.”
In the UK, Chief People Officer Phillippa O'Connor tied the cuts to the UK market rather than to AI, but acknowledged that “a graduate trainee does what a senior associate did three to five years ago.” UK graduate applications to the firm rose 35% for 2026, and PwC says it still needs junior staff to build the “human skills” AI cannot yet match.[1]
Read the two PwC statements together and the tension is visible. The official line attributes cuts to the market; the operational line concedes that a trainee now delivers senior-associate output. Both can be true, and together they describe the mechanism precisely: AI has not made juniors redundant, it has compressed three to five years of apprenticeship into the first one. That is a training design problem as much as a headcount one, because the old apprenticeship, learning by doing the production work, is the part that has been automated.
The AI-native firms that skipped the pyramid
New entrants are not reshaping the pyramid; they never built one. Queen's Tower Advisory and Unity Advisory run roughly 80% AI agents to 20% human consultants, scaling on agent capacity rather than headcount. Xavier AI is building tools meant to replace consultants outright.[2] Their significance is not their size but their proof of concept: a client deliverable can be produced without the junior base the incumbents were built on.
Diamond, obelisk, hourglass: the three replacement shapes
The report identifies three main structural models emerging to replace the pyramid. Each makes a different bet about where human value sits.
The diamond
The diamond keeps a thin junior base and a small senior top but widens the middle with experienced, AI-fluent specialists. It assumes firms hire fewer generalist analysts and more mid-career technical staff, such as AI engineers. The middle becomes the engine, and the firm's leverage comes from specialist expertise applied with AI rather than from junior hours.
The obelisk
The obelisk is taller and leaner: fewer layers, smaller teams and less reliance on junior staff. It rests on three redefined roles. AI facilitators are early-career consultants who turn data into fast insight. Engagement architects are seniors who frame problems, judge AI outputs and turn them into strategies. Client leaders are seniors who build trust with executives and advise them on what comes next. These roles spread across levels, and what counts is sharper thinking delivered faster with less overhead.
The hourglass
The hourglass puts senior advisors on top, an AI platform in the middle where the production work happens, and a governance base that validates outputs and keeps them auditable. It is the most explicit of the three about where the risk sits: the narrow waist is the machine, and the base exists to check it.
| Diamond | Obelisk | Hourglass | |
|---|---|---|---|
| Where leverage comes from | AI-fluent specialists in a wide middle | Fewer layers, sharper thinking, less overhead | An AI platform doing the production work |
| Junior base | Thin; fewer generalist analysts | Reduced; juniors become AI facilitators | Thin; analysts and AI operators below the platform |
| Hiring implication | More mid-career technical staff such as AI engineers | Seniors who can frame problems and judge AI output | Senior advisors plus a governance and validation layer |
| Main risk it addresses | Losing expertise depth as juniors shrink | Overhead and slow decision cycles | Unverified AI output reaching clients |
What changes for recruitment and skills
Different as they are, the three shapes reward the same two things: technical skills and governance skills. The diamond hires AI engineers into the middle. The obelisk asks engagement architects to judge AI output. The hourglass builds an entire tier whose job is validation and auditability. In every model the scarce capability is no longer producing the first draft; it is knowing whether the draft is right, and being accountable for it.
For recruitment, this inverts the traditional funnel. Firms used to hire volume at the bottom and select for judgment on the way up, over years of supervised production work. If the production work is automated, that selection mechanism goes with it. Firms will need to hire for judgment and technical fluency earlier, and teach verification explicitly rather than assume it is absorbed through repetition. PwC's observation that it still needs juniors for “human skills” is the same point from the other side: the thing juniors are now for is the thing the old model taught last.
What this means for firms and for consultants
For firms, the choice of shape is downstream of a choice about business model: a firm selling outcomes needs a different structure from one selling reusable assets or premium handmade advice. That link is developed in how AI is changing consulting business models and pricing. For individual consultants, the message is more direct. The skills that made a strong analyst in 2022, speed and polish in production, are the skills AI has commoditised. The skills that make a strong consultant in 2026 are the ones every replacement shape pays for: framing the problem, directing the tools, verifying the answer, and owning the recommendation in front of a client.
This analysis expands chapter 2[3] of The State of AI in Consulting 2026.
Notes and sources
- PwC statements on headcount (2024 to 2025), the withdrawn 100,000-hire target, AI training of 315,000+ staff, and UK graduate applications; remarks by Mohamed Kande (Global Chair) and Phillippa O'Connor (UK Chief People Officer)
- Queen's Tower Advisory, Unity Advisory and Xavier AI, public descriptions of their operating models
- Spaik, The State of AI in Consulting 2026, chapter 2
Figures attributed to third parties are their own reported data; Spaik did not produce those statistics. Where the text offers an interpretation, it is Spaik's own.
Continue reading
- The State of AI in Consulting 2026The full report this analysis is drawn fromRead
- How AI is changing consulting business models and pricingThe economics behind the new shapesRead
- AI agents in consultingThe tools taking over analyst workRead
- AI training for consultantsBuilding the judgment the new structures rewardRead
Working with Spaik
Fewer producers, more people who can judge the output
Every replacement for the pyramid rewards the same things: technical fluency, verification, and the judgment to direct AI rather than compete with it. Spaik builds those skills in consulting teams, from onboarding cohorts to partner sessions, on the firm's own tools.