AI isn't going to kill consulting. It's going to revitalize it.

A few weeks ago, I asked Claude a question: “Give me a sector benchmark analysis on the digital maturity of asset management firms in Hong Kong.” Within four minutes, I had forty pages of well-structured, properly cited content, complete with recommendations.

This work would have kept a junior consultant busy for a week. And honestly? The result was usable and actionable.

I run a consulting firm specializing in AI transformation. I could have found this threatening. Instead, I saw it as confirmation of a belief I’ve held for two years: the disruption of our sector is not a threat to consulting.

It is a death sentence for generic consulting.

What is already happening

The numbers don’t lie. Tools like Lilli at McKinsey or Deckster at BCG can perform about 80% of a junior analyst’s typical work (research, synthesis, slide generation) in a matter of seconds. McKinsey has deployed thousands of AI systems alongside its consultants. A Harvard/BCG study documents that consultants using AI complete their tasks 25% faster with 40% higher quality.

This is not marginal optimization. It is a structural overhaul of the consulting production model.

The traditional pyramid model (a broad base of analysts absorbing deliverables, middle managers coordinating, a few partners at the top) is directly threatened. What a team of six used to do can now be done with two senior consultants and AI agents. The big firms know this. McKinsey has said so publicly.

The dilemma to be resolved

Here’s what’s fascinating: the industry giants find themselves caught in a trap they would have diagnosed in their own clients. Technology allows them to be more efficient. But their revenue depends on their inefficiency. This explains, by the way, why “the cobbler’s children have no shoes” is a recurring reality in our sector. It is this truth that can no longer exist, and I send my regards to all the cobblers, by the way.

Charging by the hour for consultants when AI does 60% of the work: this model’s days are numbered. Clients know this. And yet, changing would mean sawing off the branch on which the entire structure of compensation, promotion, and valuation of the big firms rests.

The result: we invest in AI to appear innovative, but we deploy it cautiously so as not to disrupt the cash cow. This is the innovator’s dilemma, management consulting style.

Who will disappear and who will thrive

Disruption isn’t hitting consulting as a whole. It’s striking with surgical precision.

What is in mortal danger: generalist consulting without proprietary IP, the production model based on the volume of junior consultants, and time-based billing for deliverables that AI produces in minutes. This value proposition erodes every quarter.

What is holding up, and even gaining value: genuine sector-specific depth—knowledge of regulators, internal dynamics, and key stakeholders. Supporting human transformation—changing an organization, building buy-in, managing internal politics. Certified governance, particularly regarding AI itself, with frameworks like ISO 42001 that create a new scope of value that very few have mastered yet.

What is thriving: firms that have understood that AI is not their competitor. It is their production infrastructure.

The Four Imperatives for Survival

First imperative: stop selling time, start selling results. Performance-based billing is no longer a cutting-edge option. It is the only way forward. Clients who see AI at work will no longer pay the same price for the same hours. McKinsey has understood this: approximately 25% of its fees are already tied to performance metrics.

Second imperative: build a defensible proprietary IP. Not slides. Not generic frameworks. A structured methodology, rooted in years of industry-specific experience, embodied in a body of data and a logic that AI can amplify, but not replicate. Value is shifting from the execution layer to the methodological layer.

Third imperative: specialize. The data is clear: consultants specializing in specific industries or functions command premiums of 30 to 40% compared to generalists. Sector-specific and regulatory depth is precisely what AI masters least well in the short term.

Fourth imperative: agentify your offering before your clients do it themselves.

The firm that arrives with an agent configured for French banking regulations, trained on ten years of real-world engagements, is no longer playing on the same field. It sells access to a platform.

What this means for us

At Gabriel Greenfield, we made a strategic choice two years ago: to build Meridian as a proprietary methodology, not as a traditional consulting firm.

Five frameworks, six transformation phases, four-tier governance: a structured body of work that we are currently transforming into an agent-based system to deploy it as a platform.

It is no coincidence that we position ISO 42001 as a core offering. AI governance is the new perimeter of trust. Whoever masters this framework in the coming 12 to 18 months builds a competitive advantage.

The real question

AI disruption does not kill consulting. It kills undifferentiated consulting.

What will survive is the combination of deep industry expertise, structured intellectual property, certified governance, and the ability to orchestrate AI rather than be at its mercy.

The sector will polarize. On one side, AI platforms that commoditize generic analysis. On the other, firms that have built something AI cannot deliver on its own: trust, framework, methodology, and relationships.

The window to position oneself in this second category is open, but not indefinitely.