Leading through the agentic revolution

The World Economic Forum is one of those rare events where you actually get radical honesty. Global leaders and executives from the most powerful institutions gather in Davos for a week, not for viral soundbites or headlines, but for raw conversations about what comes next. Conversations that truly shape the year ahead.

AI dominated the agenda again—no surprise. What was different this year was the tone. Less frustration about why this isn’t working, and much more focus on how to actually operationalize agentic AI at scale.

And the most forward-thinking leaders are taking it one step further: “How do we actually lead through this revolution?”

Because this next phase of AI is going to reshape how work gets done, how value is created, and how people contribute for generations. You can’t just pass that off to your CIO.

I spent the week with leaders who came to Davos ready for something to radically change – on stage at Axios and The Female Quotient, in sessions with McKinsey and KPMG, in closed-door roundtables, and in the endless hotel-lobby conversations, coffee meetings, and even one impromptu meeting right on the Promenade.

Here’s what forward-thinking leaders are doing to lead the charge:

We’re entering REBUILD mode

There’s a critical gap between AI’s promise and the reality on the ground.

At McKinsey’s Tech & AI Breakfast, Kate Smaje put numbers behind what so many leaders feel: 62% of enterprises are running AI pilots. Only 10% have scaled.

I’ve sat across from dozens of CMOs and GTM leaders over the past year as they walk me through their “AI strategy.” Twenty experiments in marketing. Fifteen in finance. Twelve in operations. All producing impressive demos for the board. None actually changing how the business runs.

What’s happening is consistent: companies are using AI as a productivity layer, not as a driver of transformation. They’re scattering pilots across teams instead of rebuilding the workflows that actually move the business.

But you can’t layer AI onto broken workflows and expect transformation.

Most enterprise workflows weren’t designed for a world where execution is cheap and abundant. They were designed for a world where human effort was scarce and complexity had to be managed through an exhausting web of handoffs and approvals. Coordination was the job.

When AI gets bolted onto that system, it might make individual steps faster — but it doesn’t change the work.

The leaders pulling ahead are entering what I call “rebuild mode”. Here’s what they’re doing:

• Stripping the workflow down to the outcome. Literally put it on a whiteboard. Circle the steps that are essential and cross out the ones that exist purely out of habit. Instead of “Let’s make our 13-step approval process 10% more efficient,” ask the harder question: why do we still need 13 steps at all?
• Looking for the handoffs—the “momentum killers.” These are the places where brilliant ideas die in a spreadsheet waiting for three approvals. Where a customer request gets routed through five people before anyone takes action. Collapse ownership. Make one person accountable for the outcome end-to-end.
• Eliminating the “coordination tax.” The endless Slacks, status meetings, and “just looping you in” emails. That’s work about work. If a workflow requires 100 clicks, 20 emails, and 17 tabs just to stay aligned, it’s broken. Kill the check-ins and replace it with an agent that moves work forward automatically.

The companies that rebuild aren’t just getting more efficient. They’re becoming fundamentally more powerful because they’ve removed the organizational drag that constrained them all those years.

Radical simplicity beats org charts

For decades, we’ve rewarded the wrong kind of leadership.

We’ve promoted leaders who manage complexity well. The bigger the org chart, the larger the budget, the more layers you oversee—the higher you climb. The ability to keep a sprawling system running became the definition of success.

The leaders I sat with in Davos recognize that entire model is working against them now. In an agentic world, leadership isn’t about managing complexity, it’s about destroying it.

The people best positioned to do that aren’t sitting in your boardroom. They’re three or four levels down. The operators, the GTM leaders, who live in the workflows every day and can see exactly what needs to break.

The companies actually scaling have found these unsung heroes — and they’re everywhere, I guarantee it — and have given them real authority. They’re collapsing business-as-usual and taking a machete to the complexity that only existed because humans were the bottleneck.

After years working around broken systems, these folks don’t need much convincing. They need permission. And the enterprises that win in this next phase will be the ones who give it and get out of the way.

Jobs are no longer about execution. They’re about leverage.

One question kept surfacing: “What does a ‘job’ even mean now?”

One leader put it bluntly: “I need to tell 80,000 people what their job is going to be in six months. What do I say?”

That question doesn’t come out of nowhere. For decades, we’ve structured jobs around productivity: how many tasks you complete, how fast you execute, how busy you look. We built entire career ladders around that model—move up, manage more people, control bigger budgets.

That model breaks once execution stops being scarce. When people across your organization can automate and orchestrate work with a single line of instruction, execution becomes table stakes. It’s creativity and judgment that take over.

We’ve been asking too little of our people. Value can’t be tied to doing the work anymore. The next era jobs are going to look a lot different:

• Skill stacking replaces handoffs. People will be able to solve their own problems end-to-end instead of handing off to specialists. When your social media manager can also build the automation, pull the insights, and optimize the strategy—that’s the new bar.
• The rise of “generalists.” Roles themselves start to evolve. Job descriptions shift toward full-stack ownership. A content strategist becomes a growth architect, orchestrating content, channels, analytics, and optimization as one system.
• Growth moves sideways, not up. It will move laterally. When people own outcomes end-to-end, growth stops moving up a ladder. It moves outward —taking on bigger problems, broader responsibility. One leader I talked to restructured their entire incentive model around “impact per person.” Not how big your team is, not how busy you look. But how much leverage you create.

Once leverage becomes the measure of value, who gets access — and who feels permission to use it — becomes critical.

We’re watching a transformation that rivals the industrial revolution unfold in real time—and right now, women are disproportionately landing in the DISRUPTED bucket (57% vs 43%) while men are more likely to get AUGMENTED (46% vs 54%).

One executive told me, “This is playing out on my team right now. The guys are asking for agent access and experimenting. The women are waiting to be told what to do.” That’s not about capability or skills—it’s about who feels permission to experiment.

If leaders don’t intervene, AI won’t just scale productivity — it will quietly scale inequity.

This is where leadership accountability becomes unavoidable. You can’t let AI transformation happen to your organization and hope it sorts itself out fairly. You have to deliberately design for it.

Your people are your innovation engine. Raise the ceiling on what they’re capable of.

Get your hands dirty

If there’s one takeaway, it’s this: get your hands dirty.

The leaders driving the agentic revolution inside their orgs aren’t waiting on strategy decks or perfect answers. They’re digging into what’s broken inside their organizations and owning the change.

Redesigning how people work, grow, and create value just became the job.

The organizations who move decisively on this will create something fundamentally more powerful: more capable, more resilient, and vastly more human.