Is AI dismantling McKinsey & Co's oligopoly or strengthening it?

It was around 2014, and I'd booked a mentorship chat with an up-and-coming social media agency owner I admired. We met at 54 Stanley's chic industrial vibe in Johannesburg, ordered sarmies, and I pitched my aspirational idea: leverage my broadcasting career, business undergrad training, and modest entrepreneurial experience into a pivot towards top-tier independent consulting.

By the end of the meeting, I felt deflated and embarrassed. He seemed almost gleeful as he pulled no punches. But to be fair, he raised valid points I needed to hear about the nature and practice of top-tier consulting. Apparently, even running a successful social media agency—itself starting to challenge certain advisory work once reserved for management consultants—put him nowhere near the clout necessary to play at the level I envisaged.

He cited elite branding and reputational considerations, privileged network effects leveraging everything from old money relationships and generational wealth to private school connections and corporate old boys' clubs. Then there was the technical execution gap: my research depth, executional speed, and polish would be hopelessly inadequate. Assuming, of course, I even got the opportunity to pitch to the right parties at the right level at all.

The advice? Save myself the embarrassment. Stay in my media lane. Maybe just find a real job. Not that he was offering one.

The oligopoly adapts

In October 2025, OpenAI awarded McKinsey for surpassing 100 billion tokens in platform usage. That translates to roughly 75 million pages of text processed through their internal AI chatbot "Lilli," trained on 100 years of the firm's work. The firm that built an empire on bespoke analysis is running insights through the same underlying technology available to anyone with internet access.

Ah, but there’s more to it. McKinsey is being disrupted as much as it's forcefully evolving. They've deployed the tool across 70% of their 45 000 employees and are simultaneously laying off staff whilst investing heavily in AI integration. According to their own 2025 global AI survey, 78% of organisations are now using AI. The big consultancies aren't twiddling thumbs in the face of disruption, but rather racing to rebuild their models around the very technology threatening to commoditise their core offering.

Meanwhile, Kenya leads the world in ChatGPT usage at 42.1% of internet users, ahead of the UAE (42.0%), Israel (41.4%), and miles ahead of Japan (5.8%), China (7.3%), and the United States (19.1%). The analytical horsepower once concentrated in Sandton, Canary Wharf, and McKinsey's New York headquarters is now globally distributed, though data costs and infrastructure gaps mean this democratisation remains uneven.

A new generation of AI-first agencies and consultancies is emerging with explicit intent to take work off McKinsey's plate. These aren't just traditional firms adding AI capabilities—they're AI-native operations built from the ground up to deliver comparable outputs at fractions of the cost and time. When campaigns that typically required 14 people over 45 days can now be delivered by two people in 12 days at 80% lower cost, the status quo fractures.

Beyond structured firms, solopreneurs and hybrid independents are redefining what freelance consultancy looks like. This isn't the gig economy's ride-sharing tier. This is high-end strategic work like fractional C-suite roles, ecosystem strategy, and stakeholder engagement, becoming the aspirational layer within the broader democratisation of professional services. With projections suggesting 1.5 billion people globally engaged in gig work and half of the developed world's workforce participating by 2027, what was once exclusively the preserve of elite consulting firms is becoming viable for independents with the right stack.

Where the real advantage lives

Oracle founder and executive chairman Larry Ellison outlined his "two phases of AI" thesis at Oracle AI World 2025: Phase one (current) trains models on public internet data. Phase two (more valuable) involves AI reasoning with proprietary enterprise data.

"Your competitive advantage doesn't come from public data," Ellison said. "It comes from your proprietary customer data, your operational data, your industry-specific knowledge."

This is the structural problem for traditional consultancies, though not an insurmountable one. They don't own that data, but firms like Deloitte are building proprietary AI on client data partnerships, blurring the lines between vendor and advisor. Still, boardrooms are asking different questions. Not "Can you analyse this for us?" but "How do we extract strategic value from what we already possess?"

This shift opens space for a different value proposition. Independent consultants and smaller firms aren't competing on case study libraries or proprietary frameworks. They're competing on genuine connectivity, contextual fluency, and the ability to activate networks for commercial outcomes. When organisations engage independents now, the brief isn't "analyse our market position." It's "help us understand the ecosystem dynamics we're missing and position our engagement strategy for traction."

The uncomfortable truth

Only 3% of AI users currently pay for premium subscriptions, yet adoption is exploding. Even industry leader ChatGPT converts just 5% of weekly active users to paying subscribers. This reality exposes everyone. The same tools dismantling parts of McKinsey's moat are available to independents trying to build sustainable practices.

If clients can run scenario analysis through ChatGPT, access ecosystem intelligence through networked platforms, and synthesise strategic options using Claude, what exactly is anyone selling beyond facilitation, context, and connectivity?

Yet the infrastructure justifying credentialism hasn't simply crumbled; it's been rebuilt differently. Boardrooms still need people who understand how ecosystems actually function, who can navigate gaps between expectations and realities, who bring genuine network effects rather than contact lists. But as Deloitte's AI trends note, adoption barriers like ethics, compliance, and skills gaps create friction that benefits those who can navigate complexity—whether big firm or sophisticated independent.

There's a structural shift here reminiscent of broadcasting's disruption. YouTube eclipsed traditional television. Social media breaks news. Modern influence leverages niche, context-specific stacks enabled by digital platforms. Now layer AI onto that foundation, and what seemed impossible a decade ago becomes achievable—though not without new risks, including AI potentially exacerbating inequality in gig work by displacing millions of jobs whilst creating new categories that favour those with existing advantages.

What's actually changing

The consulting oligopoly isn't disappearing. McKinsey, Deloitte, and the others have deep client relationships, regulatory expertise, preferential partnerships with Big AI and brand equity that won't evaporate. But three things are shifting:

• The premium they can charge for certain work is shrinking. Junior consultant tasks are being automated. And barriers to entry have lowered for independents with sophisticated AI fluency.
• More importantly, the nature of strategic value is changing. It's moving from analysis and synthesis (which AI handles increasingly well) toward connectivity, context, and the ability to activate networks for commercial outcomes. Execution, baby! 
• The consulting model always relied on information asymmetry. AI doesn't eliminate the need for judgement, but it dramatically reduces the premium chargeable for synthesis alone. And it places a premium on finessing real-world outcomes, not just advising how to get it done.

The new question

When will the adage "no one ever got fired for hiring McKinsey" give way to viewing them as potential waste? When does bulletproof strategy become properly harnessing what AI-wielding independents and smaller firms can deliver?

The social media (now marketing) agency owner who advised me a decade ago raised genuinely valid points about how things worked then. His assessment was perfectly rational within the rules of that moment. What neither of us anticipated was how quickly those rules would change, and how the big firms would adapt whilst the door opened for others. The question now is whether any of us can stay standing once we've stepped through.

Andile Masuku is Co-founder and Executive Producer at African Tech Roundup. Connect and engage with Andile on X (@MasukuAndile) and via LinkedIn.

*** The views expressed here do not necessarily represent those of Independent Media or IOL.

BUSINESS REPORT