AI adoption cannot be top down—here’s how this HR leader is making organizations go AI native

Faye Wai

Last updated on Oct 7, 2025

QUICK SUMMARY

Johannes Sundlo has been leading HR teams for 15 years. And now, he's helping organizations implement AI successfully. The operative term, here, is "successfully". There's a big difference between successful AI implementation and what 90% of companies are doing.

TABLE OF CONTENTS Successful AI Integration AI in Performance Reviews & Recruitment Overcoming AI Adoption Barriers AI-native Organizations AI Tool Stack The Rise of AI managers

Here's what Johannes had to say about doing AI right.

From military service to HR leadership

I'd like to think that leadership is the most crucial part of what shapes an organization.

I started my leadership journey in the Swedish military. In Sweden, everyone has to do military service. I did mine for 15 months — it was my first time in leadership, and it sparked my curiosity.

Eventually, that curiosity led me to HR, and that's what I've been doing for the last 15 years, in one way or another. Right now, I'm a consultant helping organizations and leaders understand the ins and outs of AI.

How AI is reshaping leadership and team management

Is leadership changing in an AI-first world? Both yes and no.

On one hand, work is still about delegating problem-solving tasks to different people.

On the other hand, employees now have new capabilities that let them work more effectively. They can solve more tasks, be more creative, be more innovative, free up time, and, in theory, take on more work. That requires a new mindset when it comes to leadership.

Leadership is becoming less about, “Who do I give this problem to?” and more about, “What type of competence do I need for this, and where can I find it?”

Leadership is becoming less about, “Who do I give this problem to?” and more about, “What type of competence do I need for this, and where can I find it?

Johannes Sundlo

AI Adoption Consultant

In the past, organizations scaled by hiring more people to take on more work.

But that equation is changing. AI lets fewer people do more, and in some cases, AI itself can handle part of the work.

That shift demands a new lens on leadership. One where leaders must constantly re-evaluate how problems are best solved and by whom — or by what.

The turning point: When AI transformed my leadership approach

In the beginning, I thought AI would be a side tool that I used every now and then. But within weeks, I caught myself defaulting to it for things I’d never planned — drafting outlines, checking assumptions, and even sense-checking decisions.

I remember the first time I asked GPT-5 to digest a stack of heavy AI reports for me.

Normally, I’d block off a weekend with highlighters and coffee, trying to make sense of trends and insights. Instead, I fed them in, asked for a consolidated view, and it came back with a structured analysis, clear themes, and even comparisons across sources.

That changed something for me. Not just because it saved time, but because I started to think differently about my own role.

My job isn’t to read everything anymore; it’s to ask sharper questions and then interpret what comes back. That’s a leadership shift from being the one who processes all the material to being the one who frames the problems well enough for AI to help solve them.

My job isn’t to read everything anymore; it’s to ask sharper questions and then interpret what comes back. That’s a leadership shift from being the one who processes all the material to being the one who frames the problems well enough for AI to help solve them.

That shift affected my team, too. People became more open to trying and less afraid of “not being experts.” Once they saw that AI could take on the grunt work, they started thinking more creatively about what was possible.

So, if I had to point to one moment that changed the way I lead, it was when I realized that I don’t need to prove my value by consuming all the information myself. I need to lead by defining what matters, and let AI handle the rest.

Essential competencies for successful AI integration

I mentioned earlier that leaders have to identify necessary competencies for AI integration. Doing that is both simple and hard at the same time.

Simple, because we usually know the basics: You need technical skills, data literacy, and the ability to work with vendors and tools.

Hard, because AI cuts across roles, and what’s “necessary” isn’t just coding or prompt engineering.

Here's my process:

I start with the work itself. What are people actually trying to achieve?
Then, I look at the friction points. Where is time wasted? Where are decisions slow? And where do we get inconsistent quality? That’s where AI might fit, and that’s also where competencies need to be built.

So, sometimes it’s about teaching managers to ask better questions. Sometimes it’s about giving HR people the confidence to experiment with tools. And sometimes, it’s about having someone on the team who can connect the dots between IT, compliance, and the actual end-users.

In other words, I don’t look at a fixed list of skills. I look at the organization and say: "What do we need to move forward without tripping over our own feet?" That usually gives me a much clearer picture than any framework on paper.

How AI improves performance reviews and recruitment

Two areas I’ve really changed with AI are performance reviews and recruitment.

Performance reviews

In performance reviews, AI helps me surface patterns in feedback. Instead of managers spending hours trying to piece together comments, AI pulls themes, highlights strengths, and even finds blind spots. That means conversations can focus less on “what happened” and more on “what do we do next.” It’s made reviews faster and, honestly, more useful.

Here's what that looks like in practice. Let's say you have a running doc with your employee — preferably including transcribed meeting notes. Ask a GenAI tool to summarize the last six months. And then you can, with the help of AI, see what's been going on in the last six months — what you talked about, what's been going well, what's been going poorly, etc. It's also a great tool for giving you feedback about yourself.

Recruitment

In recruitment, I’ve used AI to draft job ads, screen applications for fit, and even reframe candidate feedback for hiring managers. The big win isn’t automation for its own sake; it’s consistency. Everyone gets the same structured information, which reduces bias and speeds up the process.

Of course, it's important to remember that AI doesn’t replace human judgment; it just clears the noise. Managers can spend more time making decisions and less time sorting through text. That shift alone has made both recruitment and reviews more strategic, instead of purely administrative.

Why AI adoption stalls and how to overcome barriers

There's still a big gap between the promise of AI and the messy organizational reality of it.

There's this shiny idea that AI is this magic productivity engine. Most leaders talk about efficiency gains, but day to day, you still have outdated processes, compliance worries, and people who don’t know when they’re “allowed” to use the tools.

So the disconnect isn’t really about the tech. It’s about trust, clarity, and design. Organizations buy licenses and think they’re “doing AI,” but they don’t anchor it in workflows, policies, or culture. That’s why adoption stalls.

In my own leadership, I try to make it less abstract.

Instead of telling your team, "AI will save us time," try saying, "What are the three most boring tasks we do every week? Let’s see how AI can reduce those." That small shift lowers fear and makes the value visible.

And in org design, I push for clear ownership. Someone needs to bridge HR, IT, and legal so employees aren’t stuck guessing what’s okay.

So, to me, the real gap is between aspiration and permission. AI promises the future of work, but people still need someone to redesign the present so they can actually use it.

Johannes' Tip

AI promises the future of work, but people still need someone to redesign the present so they can actually use it.

Proven strategies to make AI adoption stick in organizations

Let's dig deeper into that.

For me, “AI-ready” isn’t a visual on a slide. It’s when teams actually change how they work. That’s the space I focus on: Helping organizations move from strategy into practice. Not just talking about what AI could do, but sitting with teams, redesigning their workflows, and showing them how to use AI right then and there.

I’ve done this with leadership teams, HR teams, finance teams — you name it — across different industries, and the pattern is the same: People don’t need more theory, they need to see how AI connects to their reality.

Once they experience that, adoption takes off.

The problem is that most organizations treat AI like another IT rollout: Buy the licenses, send an email, and expect people to adapt. That never works.

You have to grow confidence, competence, and culture so people feel ready to actually use AI in their jobs. That’s what makes an org truly AI-ready.

Focus on the practical side. Sitting with teams, redesigning processes, and showing what it looks like in their real work. When people experience that, the shift sticks — especially if you can stop work from being boring. Few can resist eliminating boring work.

Focus on the practical side. Sitting with teams, redesigning processes, and showing what it looks like in their real work. When people experience that, the shift sticks. Without it, AI just becomes another unused tool in the stack.

Johannes Sundlo

With that, you change the discussion and your team's feelings toward AI. And slowly and steadily, it transforms the culture.

Without it, AI just becomes another unused tool in the stack.

Why experimentation is key to AI adoption

Your team doesn’t need perfect training to start. Once they see one real use case that saves them time, they will run with it. Adoption spreads more through curiosity and small wins than through any top-down rollout.

That goes for leaders, too. If I could give you any advice, it's this: Stop waiting for perfect clarity. The tech is moving too fast for that. Instead, pick a few areas where you can experiment safely and start building muscle.

Stop waiting for perfect clarity. The tech is moving too fast for that. Instead, pick a few areas where you can experiment safely and start building muscle.

Stop waiting for perfect clarity. The tech is moving too fast for that. Instead, pick a few areas where you can experiment safely and start building muscle.

Leaders often think they need a full strategy before they act, but in this space, learning comes from doing. Your job is to create the conditions for people to try, fail, and learn without fear. That’s how literacy and adoption grow.

And remember: Don’t confuse buying tools with leading transformation. The real work is cultural — giving permission, setting guardrails, and showing by example that AI is part of how you operate.

If you do that, you’ll be ahead of 90% of organizations still stuck in the PowerPoint stage.

What it means to become a truly AI-native organization

Right now, most organizations aren’t running truly agentic systems. What I see more of is automation, chaining tools together, smoothing workflows, and reducing manual steps. That’s the starting point.

Where it gets interesting is with businesses that are redesigning from the ground up as AI-native. That’s a very different playbook. And that’s exactly the space I work in now: Helping companies rethink their processes, structures, and even business models so AI isn’t an add-on, it’s part of the design.

Here's an example. One of my clients rebuilt their entire hiring process from the ground up. Instead of recruiters spending hours screening, we added AI that creates shortlists and drafts outreach. Recruiters then spend their time on candidate relationships and strategy.

It’s the same goal, but the process and roles look different when AI is baked into the design.

Inside my AI tool stack: Everyday use and experiments

Right now, my stack is a bit of a mix: Core tools I rely on every day, plus a few I’m experimenting with.

The ones I use the most are ChatGPT Pro, Claude, and Gemini. I switch between them depending on the type of task.

ChatGPT for my everyday work, as well as deep research.
Claude for everything that has to do with coding and visualizations, such as creating dashboards for KPIs.
And I have Gemini embedded in my Google products. Plus, I love the new image generator, Nano Banana!

For automation, I lean on n8n and Zapier.

For building and experimenting, I’ve been using Lovable and Replit. I've built my own ATS, my own Employee Engagement survey, and now I'm experimenting with a manager portal.

And for content, I work a lot with ElevenLabs for voice and Gamma for slides.

The big unlock: Chaining AI tools for end-to-end workflows

The big shift here is that I’m moving toward chaining tools together. Instead of asking, “Which tool is best?” I’m asking, “How do I connect these tools so the workflow runs end to end without me stepping in?”

That’s been the big unlock recently.

For example, for emails, I've connected my Gmail to a ChatGPT agent that determines whether it's a marketing email or an email that needs my reply. If the latter, it will also draft a response. Having trained on my data, it knows my tonality, etc. That alone was a game changer for me.

The most impactful AI tool in my workflow

If I had to pick one tool, it’s GPT-5 Pro and its data capabilities. Like I said, I’ve thrown huge datasets, heavy reports, and long documents at it, and it just handles them.

The impact is twofold. First, speed. I can go from raw material to a structured analysis in hours instead of days. Second, perspective. It doesn’t just summarize; it surfaces patterns I might miss if I were buried in the details. That changes how I prepare for clients and how I make decisions.

So yes, I’m a bit obsessed. It’s the first time I’ve felt like the bottleneck isn’t the data or the reading time; it’s just how good a question I can ask.

The rise of AI managers and the future of leadership

This is all just the beginning. In the next five years, we'll start seeing AI managers.

In some ways, we already do. Think about Uber — no one is manually assigning you to a driver; the system makes that call. That’s management. Tasks are allocated, performance is tracked, and resources are optimized, all without a human manager in the loop.

I think we’ll see that logic move into more knowledge work. Not just scheduling rides, but assigning projects, balancing workloads, even handling parts of performance management. Humans will still lead, but AI will take over big chunks of the coordination and decision-making.

That shift will force leaders to redefine their roles once again. And it'll be less about “who does what” and more about “why are we doing this and how do we make it meaningful.” In other words, leadership will become less operational and more about purpose, trust, and context.

Learn more

You can learn more from Johannes on his newsletter, FullStack HR, where he discusses the future of work and leadership.