McKinsey Has 20,000 AI Agents. Now They're Grading You On Working With Them.

Most coaches think AI fluency means knowing which tools to use.

McKinsey just defined it as something else entirely.

Not tool selection.

Not prompt engineering.

Judgment.

The ability to take AI output... challenge it... contextualize it... and turn it into something a real client can act on.

If you're coaching executives, leaders, or consultants... this matters more than any AI tool you'll recommend this year.

What McKinsey Actually Did

The numbers first.

That's not incremental growth.

That's a structural shift in how one of the world's most prestigious firms operates.

For every 2 humans at McKinsey... there are now 1 AI agents running in the background.

But here's the part most people missed.

McKinsey isn't slowing down hiring. They're actually boosting graduate intake by 12% in 2026.

They're not replacing humans with agents.

They're testing whether their humans are good enough to lead them.

The New Interview That Changes Everything

Final round candidates at McKinsey now face three separate interviews.

The last one involves Lilli.

McKinsey's proprietary AI platform. Real client scenarios. And a simple instruction:

Prompt Lilli. Review its output. Exercise your judgment. Produce a structured answer.

The assessors aren't looking for AI expertise.

They're looking for:

• Curiosity in how you approach the AI's output
• Skepticism about what it gets wrong or misses
• Judgment in adapting its answer to the specific client context
• Communication in presenting the result clearly

Those four things?

That's not an AI skill.

That's a thinking skill that now has to be demonstrated alongside AI.

The Part That Matters for Coaches

Here's the uncomfortable truth.

Your clients are being evaluated on this right now.

Not just by McKinsey. By every firm, every board, every client that's watching how leaders navigate AI.

The executives you coach are walking into rooms where AI output is on the table.

And the question isn't "did you use AI?"

It's "what did you do with it?"

AI doesn't replace good judgment. But the absence of good judgment becomes impossible to hide when AI is doing the work beside you.

Most coaching programs aren't preparing clients for this.

They're still teaching strategy, communication, leadership presence.

All essential.

But there's a layer missing.

The Three Layers of What McKinsey Is Actually Testing

Think about intelligence in the age of AI as three distinct layers.

Most professionals are being trained on the floor.

McKinsey is now hiring at the middle layer.

The coaches who win the next era are building the ceiling.

What This Means For Your Practice

If your clients are senior professionals, executives, or consultants...

AI judgment is now a competency you need to be coaching for.

The shift isn't subtle.

It's structural.

And McKinsey just made it visible.

Your Move

Ask yourself one honest question.

When your clients bring AI-assisted work into a room... are they walking in with confidence or anxiety?

If the answer is anxiety...

That's where coaching is needed most right now.

Not "how to use ChatGPT."

How to think clearly when AI is doing the analysis beside you.

How to trust your judgment when the output sounds authoritative.

How to be unmistakably human in a room full of intelligent machines.

That's the work.

And it's yours to do.