How AI Is Changing Leadership: What 30 Senior Leaders Got Wrong About Adoption

At a recent closed-door summit of 30 senior HR and people leaders from organizations including Netflix, Heineken, and TD SYNNEX, every attendee was asked to place their company on a four-stage AI maturity curve: Experimenting, Integrating, Embedding, Differentiating. Not a single one placed themselves at the top stage. Almost everyone landed between stage one and stage two.

That’s not a technology problem. The tools are ready, cheap, and everywhere. What isn’t ready is everything built around them — performance models, talent structures, and leadership behaviors still designed for a world where value was measured in output and execution. That world hasn’t gone anywhere. It’s just no longer the whole picture.

Here’s what the data from that room — and the broader research on AI adoption — actually says about why AI transformation stalls, and what closes the gap.

Why AI Adoption Stalls Even When the Tech Is Ready

This is the part most AI rollout plans get backwards. Leadership teams tend to treat AI adoption as a licensing problem — buy the tool, run a training, track logins. But usage rates and actual capability are two very different things.

The organizations that are furthest along aren’t the ones with the highest adoption scores. They’re the ones where three specific things are true: the performance model has been rebuilt around judgment and outcomes instead of raw output, someone senior actually owns the AI agenda instead of it living nowhere in particular, and leadership is visibly, personally on the same learning curve they’re asking everyone else to climb.

That last point is the one companies underestimate most. A leader who endorses transformation from the sidelines — sponsoring it, funding it, talking about it in town halls, but never being seen struggling with it themselves — sends an unmistakable signal: this change is for everyone else. Teams read that signal fast, and it’s one of the most reliable ways to stall adoption before it starts.

Trust Is the Infrastructure — Not the Technology

If there’s one theme that shows up across every serious study of AI transformation, it’s this: culture is what makes speed possible, not the other way around.

Netflix is a useful (if extreme) example. It runs without formal performance ratings or improvement plans, built instead on a simple test managers apply to their own teams: would I genuinely fight to keep this person? That level of trust and transparency isn’t a nice-to-have HR value — it’s what lets people experiment with new tools and new ways of working without waiting for permission from three layers up. High-trust cultures don’t need as much process because the process was never really what was holding performance together.

There’s a sharper data point behind this, too. One large organization rolled out a digital learning platform with 18,000 courses alongside a coaching program.

Take-up rate comparison: digital learning platform vs. coaching program (same organization)

Program	Take-up rate
Digital learning platform (18,000 courses)	30%
Coaching program	95%

The content was arguably comparable. The difference was relationship. People don’t build new capability from a course library — they build it from another person who’s paying attention to them specifically. That single stat is probably the most important finding in the entire AI-and-leadership conversation right now, and most organizations are still optimizing for the wrong half of it.

The Four Traits Separating AI-Ready Leaders From Everyone Else

Across dozens of unscripted table discussions at that same summit, four words kept surfacing independently, without anyone prompting them: curiosity, discernment, humility, and connection.

That’s a meaningful shift. For the last two decades, leadership development has largely been a skills conversation — communication skills, delegation skills, strategic-thinking skills. What’s emerging now is a character conversation. AI can generate the analysis, the first draft, the summary. What it can’t do is decide which question is even worth asking, admit uncertainty out loud, or notice when a team member needs something a dashboard won’t show. Those are exactly the muscles that character-based leadership development builds — and they’re notoriously difficult to develop from a slide deck.

The Real Gap: Individual Growth Doesn’t Automatically Travel to the Team

Here’s where most current thinking on AI and leadership stops short — and where the actual work has to start.

Coaching, done well, clearly grows individual leaders. That part of the data isn’t in question. The much harder, mostly unsolved problem is getting that growth to travel: from one senior leader who worked with a coach for six months, into the hundreds of managers underneath them who never got that same investment, and eventually into how the whole organization actually operates.

Traditional coaching models were never built to solve this. One executive gets 12 sessions over a year with a human coach. Meanwhile the 200 frontline and middle managers who are actually translating strategy into daily behavior — the people whose leadership determines whether “trust culture” is a value on a wall or something employees actually experience — get a training deck once a year, if that. This exact reach problem is explored in more depth in our research on why manager engagement is collapsing.

This is precisely the gap a hybrid human + AI coaching model is built to close. A senior leader still gets deep, human 1:1 coaching for the judgment calls that genuinely need it. But the curiosity, discernment, and difficult-conversation skills that used to stay locked inside an executive coaching relationship can now be practiced by every manager in the org, on their own time, through AI Avatar Roleplays and AI coaching that’s available at 2am the night before a hard feedback conversation, not just during a coach’s calendar slot every three weeks. That’s how individual growth actually becomes an organizational capability instead of staying trapped at the top of the org chart.

What This Means for HR and L&D Leaders Right Now

If you’re building the AI-and-leadership plan for next year, the data points to five concrete moves, not a tooling decision:

5 concrete moves for HR and L&D leaders building an AI-era leadership plan

Move	What it means in practice
1. Stop scoring adoption by usage rate	Track whether people are bringing judgment to how they use AI, not how often they log in.
2. Redesign the performance model first	A system built to reward output will actively fight an AI rollout meant to reward judgment.
3. Put senior leaders on the learning curve	Visibly, personally — not just as sponsors of the transformation.
4. Invest in trust before process	High-trust teams adopt faster and hold up better under disruption; more process rarely substitutes for either.
5. Design for reach, not just depth	The organizations ahead aren’t giving ten executives brilliant coaching — they’re getting practice to every manager who touches a team.

That fifth point is the one worth sitting with. Individual coaching has never been the bottleneck — reach has. See how Coachello’s coaching platform is built to close exactly that gap, combining ICF-certified human coaching for the leaders who need it most with AI practice that scales to every manager underneath them.

Frequently Asked Questions: How AI Is Changing Leadership

Is AI actually changing how leaders need to lead?

Yes. Research from a 2026 summit of 30 senior HR leaders found the leadership standard is shifting from skill-based to character-based, with curiosity, discernment, humility, and connection emerging as the traits that separate organizations succeeding with AI from those stalling out.

Why does AI adoption stall in large organizations?

Most commonly because performance systems, talent structures, and leadership behaviors weren’t redesigned alongside the new tools — not because the technology itself isn’t ready. Adoption stalls specifically when the performance model still rewards old-world output, no one senior owns the AI agenda, and leaders endorse transformation without visibly going through it themselves.

Does coaching scale better than e-learning for AI-era leadership skills?

In at least one documented case, a coaching program achieved a 95% take-up rate compared to 30% for a large digital learning platform with 18,000 courses — suggesting relationship-driven development outperforms content-only training for behavior change.

What’s the difference between AI coaching and traditional e-learning?

AI coaching, like AI Avatar Roleplays, is interactive and personalized — you practice an actual conversation and get feedback — versus e-learning, which is typically passive content consumption with no practice component.

Why doesn’t individual leadership coaching automatically improve the whole organization?

Traditional coaching models were never built to scale. One executive typically gets around 12 sessions a year with a human coach, while the hundreds of frontline and middle managers actually translating strategy into daily behavior get a training deck once a year, if that. Individual growth stays trapped at the top of the org chart unless it’s paired with a scalable practice layer that reaches every manager.

Ready to Turn Individual Leadership Growth Into an Organizational Capability?

Discover how Coachello combines ICF-certified human coaching with AI Avatar Roleplays to give every manager — not just your top ten — the practice to lead well in an AI-shaped world.

👉 Book a free demo or explore the Coachello platform.