If you’re leading in 2026, you can feel it: AI is no longer a future topic or an innovation lab experiment. It has arrived in the operational core of organizations — and with that, it’s amplifying a problem many companies already had: culture is struggling to keep up.

This is the key misunderstanding I want to challenge: AI is not primarily a technology story. In organizations, it’s a people and culture story.

AI has arrived — but scaling is the real challenge

There’s no shortage of experimentation right now. In fact, a PwC AI agent survey showed 65% of organizations were piloting agent systems last year — roughly two out of three. And McKinsey’s “State of AI” trendline shows how quickly adoption has moved from niche to mainstream over the last decade.

And yet, we all recognize the second part of the reality:

Many AI projects don’t scale. They remain small pilots, isolated experiments, or point solutions that never become enterprise capability. Why? Because scaling AI isn’t just about funding tools. It requires people to change how they work — and cultures to change how they learn.

The core tension: AI pulls culture apart

AI brings enormous capability to individuals: speed, content creation, research, rapid testing, faster iterations.

But this capability introduces divergent forces inside organizations:

• Speed increases — alignment decreases.
  When individuals can move quickly, coordination, shared context, and shared direction become harder. Systems drift from human intent.

• Innovation accelerates — trust erodes.
  Faster experimentation means choices diverge. Ambiguity increases. Without shared judgment and norms, teams lose the feeling of community and reliability.

This is not about AI “destroying” culture. It’s about AI magnifying the conflicts that already exist — between individual autonomy and organizational coherence, between execution speed and shared meaning.

The result is what I call: culture gets pulled apart.

Cultural drift: when stress makes organizations revert

Under pressure, organizations often revert to old patterns — even after periods of growth, innovation, and positive change. This phenomenon is known as cultural drift.

We saw it after the pandemic: many organizations proved they could work digitally and lead remotely — and then gradually slipped back into old habits.

Now AI acceleration is creating a similar stress environment. It’s fast, intense, and uncertain. People reach for something familiar — because the “new” isn’t stable enough yet to hold onto.

You can recognize cultural drift in everyday signals:

• people stop asking questions

• meetings are attended, but participation collapses

• fear replaces excitement about the new

• leaders start “managing” people more than working with them

• bureaucracy rises: rules and processes become safer than learning

• “this is how we’ve always done it” becomes the default

Raymond captured the human side of this perfectly in the conversation: when boundaries fade and expectations change, people step into the unknown — and fear of the unknown replaces curiosity.

Why drift kills performance: the domino effect

Culture isn’t the purpose of a business — performance is. Culture either enables or hinders that purpose. Cultural drift erodes performance in a predictable chain reaction:

1. Fear increases
   People choose safety over experimentation.

2. Habits take over
   New tools and new capabilities exist — but people revert to “the way we’ve always done it.”

3. Engagement declines
   Input drops. Quiet quitting, presenteeism, absenteeism intensify.

4. Curiosity disappears
   People don’t explore. They don’t learn. They don’t improve.

And here’s the critical line for 2026:

When curiosity ends, learning ends.

That’s why culture is the bottleneck. Not because learning is a nice-to-have — but because learning has become a prerequisite for doing the job.

If learning isn’t embedded in the flow of work, organizations can’t respond to changing customer needs, new technology competitors, or the speed of the market.

The common mistake: “culture campaigns” instead of behavior change

When leaders respond to culture problems, they often launch culture initiatives — and in most companies, these are primarily communication initiatives.

There’s even research showing that many leaders equate culture work with communication.

But communication alone doesn’t change culture.

A statistic I shared in the webinar is a wake-up call: 72% of formal culture initiatives show no meaningful improvement in employee trust, engagement, or retention.

Why do they fail?

Because culture is not what we say.
Culture is what we do — repeatedly — in daily work.

The shift that works: culture change is maintenance

If culture is daily behavior, then the solution isn’t a poster, a policy, or a t-shirt.

The solution is maintenance: creating the conditions where desired behaviors are easy, safe, and normal.

Not “talking about mentoring,” but mentoring.
Not “talking about exploration,” but exploring.
Not “talking about feedback,” but making feedback part of daily work.

Raymond made an important point here: continuous learning doesn’t happen through occasional trainings — it happens when leaders make learning and feedback a frequent, low-threshold part of one-on-ones and teamwork.

One practical intervention: reframe mentoring as infrastructure

My closing proposal in the webinar is simple:

Stop treating mentoring as a program for a few. Start treating it as infrastructure for culture.

When mentoring becomes accessible at scale — available day-to-day — it helps organizations:

• anchor judgment in ambiguity (you can consult expertise instead of guessing alone)

• build trust and community (people feel needed, connected, supported)

• create psychological safety under pressure (you’re not left alone in first-time situations)

• strengthen inclusion (easier access to advice increases cross-boundary connection)

But this doesn’t happen by “installing mentoring.”

To make mentoring an infrastructure, you need several elements in place:

• Framework: goals, structure, guidance (e.g., standard booking times like 15/30/60 minutes)

• Community building: small peer groups, thematic circles, Q&A formats, purposeful exchange
  (We shared a simple example from Mentessa: a weekly “Demo Day” where everyone shares progress and obstacles.)

• Structured matching: without structure, extremes emerge (some get overloaded, others never get access)

• Clarity and training: e.g., how to give feedback that works

• Feedback loops & impact assessment: “If we can’t measure it, we can’t manage it.”
  That’s where technology helps — not as the point, but as the enabler.

The takeaway

AI accelerates work — and it magnifies cultural tension. Under stress, culture drifts. Drift kills curiosity. And without curiosity, learning collapses.

That’s why culture is the bottleneck in 2026.

So the question isn’t whether you have an AI strategy.
It’s whether your organization has the daily behavioral infrastructure to learn, stay aligned, and trust each other while everything accelerates.