AI Won’t Fix Workplace Training but It Might Make It Worse

I feel a little sheepish admitting this, but I got a tattoo because of a workplace training.

It wasn’t AI enabled and there was no adaptive learning module. There wasn’t even a slide deck. Just 15 people in a windowless government training center for four days, exploring how power shows up in group dynamics. It was awkward, real, transformative.

That week literally left a mark on me. Its lessons in emotional intelligence changed how I lead and show up with others, and the wheat stalk on my arm reminds me that discomfort can be a source of growth. And yet, today, these kinds of emotional learning experiences are vanishing from the workplace. In their place? AI.

Solving the Wrong Problem

At a recent AI-in-the-workplace event I spoke at, sponsors shared a dizzying array of tools promising to make training faster, cheaper, and more scalable. But I left more worried than excited.

Yes, AI can help us learn about things: rules, systems, ideas. I get excited about tools that make cybersecurity or compliance trainings less painful or provide reminders and coaching so training stays fresh.

But we’re not creating space to learn about ourselves or each other: recognizing our patterns, offering feedback with care, or navigating discomfort together. Instead we’re zoning out during one-way webinars and hoping AI will fix workplace culture.

One example: I was once part of a team-building training where we anonymously rated each other on how open to feedback or collaborative we were. The AI system then assigned each person a training module based on their gaps. It was meant to be innovative, but it felt cold and disconnected. There was a mini rebellion, no one felt seen, and nothing changed.

Prioritizing AI Over EI

I’ve worked across multiple government agencies and teams, and my experience is that the biggest barrier to performance isn’t a lack of knowledge, it’s a lack of emotional intelligence. The supervisor who blames others instead of looking within. The teammate who shuts down dissent through passive aggression. The meetings where quiet, dissenting voices get drowned out.

Case in point is AI adoption itself. It’s just as much a human challenge as a technical one, requiring trust, vulnerability, and change management. And we’re not training people for that.

Earlier this year, the entire Organizational Development branch at my agency was eliminated. A small, community-led meditation group I co-facilitated called Mindful GSA was also shut down. It offered space each week to reflect, get perspective, and connect. Inefficient, we were told.

What the Learning We Need Looks Like

Each week, I help co-lead Mindful Fed, a cross-agency community grounded in mindfulness and connection. The technology is simple: a Zoom link. The method is ancient: shared silence, honest conversation, presence.

And yet, the impact is real. People show up more grounded, more compassionate, and more equipped to face the hard moments federal workers have navigated in recent months. It may not look like traditional training, but it strengthens our emotional intelligence and deepens our connection to ourselves and one another.

Even Snap (formerly Snapchat), a company not typically associated with slowing down, has an initiative called Council to help employees build trust, empathy, and understanding through facilitated group dialogue.

What More Human Training Can Look Like

If AI isn’t the solution to what’s broken in training, what is? Here are four places to start:

1. Practice the human stuff. Everyone’s talking about virtual reality, but the simulations we really need are emotional. Practicing discomfort, feedback, and turning toward empathy pays off when stress levels rise.
2. Give training a human context. As Dan Pink explains in Drive, purpose drives learning. Yet so many trainings are stripped of meaning. Even if you’re using online or AI modules for mandatory training, take time to explain why the content matters and how it connects to mission.
3. Do virtual learning together. AI can personalize content, but it can also isolate us. As Annie Murphy Paul writes in The Extended Mind, simply being in the same room can deepen learning. Even if it’s a cybersecurity refresher, reserve a room and do it as a group, and talk about the shared pain of constant password changes.
4. Check in with each other. Emotional intelligence isn’t a one-off training; it’s a muscle. Teams that reflect together regularly become more adaptable and resilient. Weekly pauses, like those at Mindful Fed or Council, create lasting change.

Workplace Dynamics Essential Reads

Do You Love Your Job?

Why Some Women Lift Others Up and Some Don't

AI in the Loop

How inefficient are family members? They ramble, they interrupt, they test our patience. But they’re also the ones who show up for us and help us feel seen, supported, and connected. Teams are no different. Building relationships takes time and care, but they’re what make work more fun and productive. And that’s what training should help us practice.

These days, we hear a lot about letting AI take the lead and keeping a "human in the loop." In the training realm, that gets it backwards. So yes, let’s keep AI in the loop for all its benefits, but let’s not forget that emotional intelligence and relationships are at the heart of meaningful work.