McKinsey facing its own disruption: how artificial intelligence is redefining strategy consulting

For nearly a century, McKinsey & Company represented the pinnacle of applied strategic knowledge. From its origins, it helped shape modern business thinking, standardized the logic of scorecards, pioneered the use of matrices and frameworks, and established its brand as a guarantor of certainty and intellectual prestige.

With his signature came a promise: analytical excellence, conceptual depth, and ability to execute.

Today, with the development of Lilli—an AI trained on more than a hundred years of institutional knowledge—McKinsey could be eroding precisely the symbolic and operational core of that promise. The change is not only technological: it can affect its structure, its culture and its value proposition.

For those who don't know it, Lilli scans more than 100,000 internal documents, interviews and presentations in seconds, identifies relevant experts and returns summaries that consultants can use as a starting point for analysis, proposals or slide decks. Today it is used by more than 70% of employees for research, modeling and writing recommendations.

This article explores whether McKinsey is unintentionally creating its own disruption and what this means for the future of consulting, human-AI collaboration, and the strategic architecture of companies.

The traditional model in crisis

For decades, the elite consulting business was based on an operational pyramid: a broad base of junior consultants, guided by knowledge frameworks and supervised by senior partners, replicated institutional learning with a logic of efficiency and scalability.

But that pyramid was much more than a productive mechanism: it functioned as a leadership school. The juniors, immersed in demanding projects and in direct contact with clients, absorbed tacit knowledge and developed judgment thanks to intensive mentoring. The growth ladder was not just hierarchical; It was epistemic.

With the arrival of Lilli and the automation of research, synthesis and knowledge transfer, a good part of that mechanism loses its purpose.

If AI accomplishes those tasks better and faster, what's left for the juniors? Furthermore: how is the senior of the future formed when instances of situated learning disappear?

Silent reduction, covert restructuring

The Financial Times (May 28, 2025) revealed that McKinsey cut its workforce from 45,000 to ~40,000 people in 18 months. He did not announce it: he resorted to performance reviews, back-office adjustments and selective layoffs. This stealth layoff (without ads) reveals an operational redesign without narrative.

The move avoids media panic and preserves the image of stability before clients, but postpones the key question: what firm does McKinsey want to be from now on?

Opacity takes its toll inside the home. Without an explanation of the direction, confidence, motivation and cohesion suffer; the collective purpose is blurred. In a historically intense and meritocratic culture, that symbolic erosion can hit as hard as a financial problem.

Worse still, the cut eliminates learning opportunities and autonomy for the remaining consultants. Without a clear epistemic ladder, the firm's identity and its ability to attract talent are at risk.

Operational efficiency or business transformation?

Reducing costs and automating processes protects margins: that's operational efficiency. Creating new ways of delivering value and redefining the signature symbol requires adaptive capacity. The first is tactics; the second, strategic.

For now, McKinsey embodies the saying in the blacksmith's house, a wooden knife: it boasts of guiding the business transformation of others, but postpones its own, using Lilli above all to accelerate internal tasks and protect margins.

If the firm prioritizes efficiency over value proposition, what real differential will its agent offer compared to the thousands already deployed by rivals, clients or startups? It could be sowing its own disruption.

The automation of past knowledge is a trap if there are no mechanisms for updating, opening and external validation. An AI trained on what worked perpetuates obsolescence; without an explicit loop of new learning the advantage erodes.

Business theory has completely changed and must evolve at high speed: methodologies, tools, concepts, approaches.

Using AI only as an assistant or to deliver the current product is to stay at the initial level: the round-trip where humans and algorithms co-create fresh conceptual frameworks is missing. Without augmented intelligence and co-creation of new methodologies, the element that will define the organizations of the future is lost. As I explain in How does AI affect work teams?

The brand promise becomes measurable

If McKinsey's differential rests on Lilli, the value stops being based on reputation and relies on verifiable results before hiring, just like when comparing two software platforms. When the client can play agent against agent, the premium of elite consultancies will last only if their outputs are consistently superior.

Contrast mechanisms will emerge—and are already being tested:

* A/B tests that parallel the recommendation of Lilli, other agents and human teams.

* Direct correlation of each proposal with the KPIs that the client later achieves.

* Mixed panels (senior consultants + AI) that audit depth and rigor.

* Spread simulators: the same challenge is solved with several agents and originality, time-to-value and risk are compared.

There are still no industry standards, but the simple possibility of this comparison has a disciplining effect: the more transparent the test, the more difficult it will be to sustain prices based only on historical brands. If Lilli does not exhibit a clear advantage—unlikely given the open alternatives—the power to set premium rates will be diluted. Even without formal frameworks, mere competition between agents will compress that differential.

Comparison with other consultancies

While McKinsey quietly restructures and automates processes, the rest of the Big 5 tries different paths — although still more declarative than transformational.

BCG: Added 1?,000 AI specialists in 2024, created BCG X—an internal product and venture factory—and already generates around 20?% of its global revenue from AI services. It reported a growth of 10?% and has had 21 years of consecutive expansion, a sign that it combines traditional consulting with the construction of its own solutions.

Accenture: Despite more than 40 tech acquisitions, Accenture's recent stock market setback reinforces the alert: in just four months the firm lost USD 60 billion in capitalization and its new gen-AI contracts were reduced by half. The market rewards technology originators—Palantir is worth six times more than it was a year ago—and punishes integrators who do not prove a differential advantage.

EY Parthenon: Has invested $1?400M in AI and launched EY.ai Factory, a network of labs where mixed EY co client teams create agents and use cases. More than 5,000 consultants already use Microsoft 365 Copilot; The Parthenon sub-brand was expanded to unify Strategy + Transactions + Transformation with an AI-centric focus.

Kearney: Signed an alliance with the Dubai?AI?Campus to set up an AI Experience Center and, together with Carnegie Mellon, launched an AI Executive Program aimed at C level. Their emphasis is on hybrid talent and predictive intelligence through platforms like Nostradamus AI.

Deloitte: Received the Google Cloud Partner of the Year award (AI, 2025) and, together with Google, promotes the Agent to Agent (A2A) protocol for agent interoperability. Launched two AI Experience Centers (Bangalore and Cairo) for customers to experience Gemini models, Vertex AI and multi-modal solutions.

These moves indicate that organizational redesign is just beginning: numerous “strategic intentions” and branding actions are shared, but few real trade-offs are made. It is true that narrative remains a crucial part of competitive advantage; but without a coherent story and visible leadership, it will not be sustained in the medium term.

BCG migrates to build + advise cells, EY sets up open laboratories, Deloitte creates multi-agent environments, Kearney mixes academia and sandboxes. None have yet jumped from the hierarchical pyramid to a network of autonomous teams augmented by AI or another form of organization: they are advanced pilots, not total reinventions.

Because of the Innovator's Dilemma—and the short-term incentives that fuel it—most prefer to optimize what already works, instead of designing what's next; exactly what happens with Lilli.

Worse still, this inertia reveals their difficulty in understanding and applying new business transformation methodologies. After a decade sowing confusion in organizations—as detailed in Elite consulting firms destroyed agility, now they are going after agents—they run the risk of repeating history: calling transformation a profitable and value-destroying drill.

Meanwhile, native AI firms such as Palantir and Aleph Alpha are launching consulting without legacy, without hierarchies and with technology that eliminates intermediaries. That's where true disruption can happen.

Symbolic architecture at risk

For decades McKinsey sold much more than analysis: it sold certainty. For a CEO, hiring her was invisible insurance that legitimized decisions before the board of directors and the market. This symbolic capital was supported by three pillars:

1. Irreplaceable human judgment. The partner put his last name and hours of his own thought into each recommendation, providing criteria that no competitor could clone.

2. Methodological exclusivity. Its frameworks, databases, and internal secret sauce were inaccessible to third parties; the firm cultivated an aura of secretive knowledge.

3. Institutional authority. The brand guaranteed that any decision supported by McKinsey would be respected by investors, press and regulators.

Lilli's emergence erodes all three foundations at once. If AI produces most of the analysis, the outside view tends to be that the partner's firm is reduced to a minor revision: it loses intellectual density. Open models and fine-tuned prompts allow rivals or even internal company teams to replicate very similar outputs in a matter of weeks, liquefying exclusivity. And by revealing that much of the work is automatic, the narrative of excellence is deactivated; authority goes from indisputable to debatable.

The resulting cascade is clear:

*Premium price. If the analysis can be replicated with an internal or low-cost agent, the McKinsey guarantee no longer justifies its price: there is no longer any scarce human input or exclusive method to support the premium.

* Talent. The magnetism of the brand weakens; The most ambitious consultants migrate to firms or startups where they see greater projection and equity.

* Differentiation. With convergent diagnostics, the firm is forced to compete on price and speed, exactly what it swore to avoid.

Is there a way out? Yes, but it requires a profound and costly reinvention that collides with short-term incentives and, often, with market and shareholder pressure. Without that leap, the McKinsey symbol runs the risk of being reduced to a hollow logo: visible, but without the authority that once sustained its margins and its power of attraction.

If you choose tactical responses, you will begin a process of decline cushioned by various actions. Between the patches it occurs to me

1. Re-hybridize the value. Fusing AI and human judgment to create truly new and useful conceptual frameworks.

2. Guarantee the result. Replace the promise based on prestige with success fee agreements or participation in upside.

3. Test it in front of everyone. Subject the partner + AI duo to simulations and open comparisons that demonstrate, with data, their superiority over any external agent.

Necessary measures, yes; but, in my opinion, insufficient and difficult to execute. Without a comprehensive transformation of the business model, these adjustments will barely slow the loss of relevance.

Implications for customers

Hiring consulting is no longer decided only by the prestige of the logo. Recent regulations and companies' own experience force buyers to gradually dig deeper into what is delivered, who produces it, and how it is validated:

* Origin of output The United Kingdom requires that every public provider declare which parts of the service it performs with AI and which with human personnel.

* Degree of automation The key account documents already detail which phases—research, synthesis, proposal—are delegated to agents to evaluate risks and quality.

* Proven quality With the entry into force of the first AI Act rules in the EU, consultancies must document evidence of model performance and transparency.

* Firm approach Guides for audit committees recommend verifying whether the consulting firm seeks only tactical efficiency or also adaptive capacity that generates new value.

* Transparency and ethics Recent global surveys show that almost half of companies have suffered consequences for inaccuracy or lack of explainability, which reinforces the demand for traceability of each algorithmic decision.

The traditional bond of blind trust is mutating towards a collaborative and auditable relationship. Whoever hires consulting today needs new questions, new metrics and a new definition of strategic value; Otherwise, you'll pay for a service based on assumptions that AI has already made obsolete.

The next frontier: new business theories and tools in live mode

The competitive advantage of the future will not come from having the most powerful AI or the greatest amount of data, but from creating conceptual frameworks that do not yet exist to manage companies in a hybrid way—AI + humans—in changing scenarios.

AI is just one of several exponential technologies; The real leap will come when they begin to converge more strongly, become cheaper and democratize, the last of the 6Ds.

The pre-AI methodologies that large consulting firms continue to use today—and that feed models like Lilli—were born for a world that no longer exists; That is why any recommendation based on these assumptions is already outdated.

The challenge is to co-create living business theories, capable of updating at the pace of changes.

In this area, hybrid teams surpass both pure platforms—preyed by the past that train—and traditional consulting—trapped in slow processes and rigid hierarchies. Whoever masters the creation and rapid obsolescence of their own frameworks will gain the necessary elasticity in an environment where knowledge expires in months.

The key question is no longer who integrates technology most efficiently, or even who generates it: it is who first redefines the logic with which companies create, capture and defend value. That career, more intellectual than technical, has just begun.

The consulting industry is being reconfigured.

The consulting industry is being reconfigured. Its greatest ally of disruption is not AI startups, but the poor internal decisions of the firms that dominated 20th century consulting.

Therein lies their contradiction: they preach transformation, but they have enormous difficulties in applying it to themselves. By automating their historical spreads without redefining their value proposition—only to ride the AI ??wave, unlock a quick revenue stream, and squeeze efficiencies—they risk dismantling the foundations that sustained their prestige and margins.

McKinsey embodies this dilemma: by delegating to Lilli the methodological memory that made her unique, she runs the risk of turning her savoir-faire into an algorithmic commodity. If you do not reimagine your strategic architecture—business model, operating model, and, above all, the creation of new conceptual frameworks—you will end up fueling the same wave that threatens to sweep you away.

The opportunity exists: to lead the next generation of business theories, co-created by hybrid teams and validated in rapid market cycles. Whoever embraces this logic will go from selling certainties of the past to designing the best approaches to the future.

In a world where algorithms produce analysis and data ages in days or weeks, the only sustainable advantage will be to reinterpret reality before anyone else, and above all, have new tools to do so. That is the new frontier of strategic consulting.