We still need consultants. Practical, hard-won experience doesn't get replaced by a model that's only read about it.

What's actually changing is the business model built around that experience: a large pyramid of relatively cheap junior labour funding a thin layer of expensive partners, priced by the hour.

That model is failing in real time, not because AI has made judgment obsolete, but because it has quietly broken the economics the entire industry was built on.

The headlines say "AI is disrupting consulting," as if the whole profession were under threat. That's too broad.

The work actually under threat is a specific layer of it: the research, synthesis and deck-production labour that junior consultants have always done. That labour was never the point. It was the funding mechanism.

Understanding that distinction, and what it means for the model rather than the profession, is where this goes next.

The Pyramid Trained People Because Training Produced Leverage

The analyst-to-partner career ladder, with its brutal "up-or-out" promotion policy, isn't a cultural quirk.

It's a rigorous economic design.

A landmark 2005 paper by economists Jonathan Levin and Steven Tadelis showed why knowledge-intensive firms like consultancies organise as partnerships: clients can't easily judge the quality of advice before they buy it, so the firm needs a structure that screens out weak performers and keeps incentives aligned. Up-or-out promotion falls directly out of that logic.

Leverage—many billable people under one partner—is what makes partner-level profits large.

A partner who does the work personally is worth far less than a partner who can deploy ten people underneath them, each billing at a lower rate against the same client relationship.

But leverage doesn't exist on its own.

You can't make partner without a bench of mid- and senior-level people who can actually do the work to a high standard, and the only way to build that bench is by developing juniors into it, year after year.

Training junior people was never a pleasant side effect of hiring them. It was the mechanism that produced the leverage the whole model runs on.

AI now performs a large share of the work that apprenticeship used to build people up through, which is exactly why training investment and leverage are starting to come apart.

The Cracks Are Already Visible

This isn't speculative. Several independent, verifiable signals are pointing in the same direction at once.

Junior hiring is contracting sharply, specifically where AI substitution is most direct. Across the UK arms of the Big Four, graduate headcounts have dropped year-on-year. Pay at the top has simultaneously gone flat, with MBA compensation at McKinsey, BCG and Bain largely plateauing after years of growth.

AI adoption inside the firms is real and heavy, not experimental. Internal AI tools are already saving consultants significant amounts of research and synthesis time. Yet the fee model underneath all of this has barely moved.

That creates a productivity paradox: a firm that fully passes AI's efficiency gains through to clients shrinks the billable-hours base its profits depend on.

Meanwhile, clients are increasingly willing to challenge consulting firms over failed engagements rather than simply absorb the cost.

Taken together, these signals suggest the traditional consulting model is beginning to bend.

What's Actually Scarce Isn't Knowledge. It's Context.

Here's the part most "AI vs. consulting" discussions miss. The frameworks, playbooks and benchmarks consulting has always sold were never that scarce.

They've been taught in every MBA programme and published in business journals for decades.

What AI changes is the marginal cost of applying generic knowledge, not its availability. What doesn't get commoditised is the proprietary layer:

• A company's real data

• Real organisational constraints

• Experience gained through transformation

• Pattern recognition developed under real organisational pressure

  
  

That knowledge is hard to codify precisely because it's earned, not read.

It's also becoming the most valuable asset organisations can build.

The Five-Part Model for What Comes Next

Put the evidence together and a consistent shape emerges: not "no more consultants," but a different business model built around five shifts.

1. Structure

Smaller, senior-weighted, multidisciplinary teams replace the standing pyramid of juniors. If AI absorbs the research and synthesis work that junior leverage used to fund, the profitable unit of delivery becomes a compact team of senior specialists, each amplified by AI, rather than a large team of generalists supervised by a thin layer of partners.

2. Assets

The durable moat shifts from any single engagement's fee to an owned, continuously improving knowledge base.

Instead of producing one-off deliverables, firms build structured records of transformation patterns, failure modes and organisational context that make AI genuinely useful on future client work.

3. Talent

A faster, redesigned apprenticeship replaces the old volume-based one.

Junior consultants spend less time producing research and presentation decks and more time supervising AI, validating outputs and developing judgement.

Future hiring increasingly rewards healthy scepticism about AI output alongside structured problem-solving skills.

4. Economics

The strategy market's current instinct is to chase pure value- or outcome-based fees. In reality, broad transformations make attribution difficult.

A more practical model blends:

• Base fees for defined work

• Shared upside tied to agreed milestones

• Optional warranty or insurance-style protection for larger engagements

  
  

Rather than betting everything on one pricing mechanism, firms balance incentives across all three.

5. Trust

Brand still matters. It functions as reputational insurance for the executive who hired the firm. However, AI also narrows the gap between what large consulting firms can offer and what smaller networks of senior independents can deliver.

That creates opportunities for new entrants to compete on trust rather than simply on price. None of this displaces the large incumbents overnight.

Their existing C-suite relationships, regulatory credibility and ability to rescue failing engagements remain genuine competitive advantages.

The more likely near-term path is bifurcation: AI-native challengers and in-house client capability absorb the commodity and mid-market work, while incumbents retreat toward, and eventually acquire their way back into, the highest-trust end of the market.

Where JustifyAI Stands

We don't think the interesting question is whether AI replaces consultants.

It's whether an organisation's operating model—how it structures teams, prices work, develops judgement and earns trust—is built for a world where the cost of applying generic expertise has collapsed and the value has moved to context, architecture and accountability.

Most enterprise AI programmes still get evaluated as a tooling decision. The firms and functions that get this right are treating AI as a business-model decision instead: a strategic and architectural question, not a procurement one.

That's the work we help organisations do: understanding where AI genuinely changes the economics of a function, evaluating what to build versus buy, and designing the governance and operating model that makes the shift durable rather than cosmetic.

Conclusion

AI isn't replacing consultants. It's reshaping the business model that consulting has relied on for decades. The future belongs to organisations that rethink how expertise is delivered, combining experienced people with AI to create leaner teams, stronger knowledge assets, modern pricing models, and more effective governance.

Success will come not from replacing human judgement, but from redesigning the operating model around it.

Sources

• Levin, J. & Tadelis, S. (2005). Profit Sharing and the Role of Professional Partnerships. Quarterly Journal of Economics.

• Scottish Financial News (2026). Big Four slash graduate jobs as AI takes over entry-level tasks.

• Fortune Education (2023). New MBA grads at McKinsey, BCG and Bain can now land base salaries of nearly $200K.

• Poets & Quants (2026). Consulting Pay: What MBAs Earned In 2025.

• Entrepreneur (2026). McKinsey Is Using AI to Create PowerPoints and Take Over Junior Employee Tasks.

• Business Insider / AI Weekly (2025). McKinsey Ties 25% of Fees to Outcomes as AI Erodes Billable Hours.

• Becker's Spine Review (2025). Zimmer Biomet sues Deloitte for $172M.