HR’s AI moment - why people leaders must own the cultural transformation

Recently, I’ve been noticing a common thread in my peers' discussions about AI. Across the enterprise, leaders are paying outsize focus on tools, models, and infrastructure. But we’re spending less time figuring out why so many of us have yet to see meaningful ROI from AI.

In large part, Atlassian’s AI Collaboration Index — CXO Insights report echoes this idea. Just 4% of companies are seeing real ROI from AI, and just 1 in 5 leaders say AI has improved their org’s innovation. Worse yet, over a third say AI has actually slowed their teams down.

So where are we going wrong?

Here’s what I can tell you as Chief People Officer — too many companies are over‑indexing on technological transformation and under‑investing in cultural transformation. AI shouldn’t be treated as a technology roll out led by the IT function alone — it must also be driven by HR leaders and people teams.

We need to shift mindsets, build trust, and shape new behaviors. And because HR sits at the nexus of all three, it’s our leaders who are best positioned to lead this change. Here’s what I’ve seen from our experience at Atlassian.

Adoption — and experimentation — trickles down from the top

One of the most important things we’ve found is that the onus is on leaders to gin up AI excitement among employees. In fact, workers who have seen their manager model AI are 4x more likely to use it themselves throughout the day.

By being open about their own AI usage, leaders position AI as a teammate — instead of a threat. Better still, they debunk the fear of AI as a human replacement by showing how it actually augments human work.

But modeling alone isn’t enough. People learn most effectively through doing. The most valuable use cases emerge when employees are empowered to experiment on their own — and share what they learn with their teammates.

I see this every day as teams experiment with our own AI teammate, Rovo. When teams apply Rovo to real use cases without fearing 'getting it wrong', they uncover the high-value opportunities that impact workflows and build enduring confidence.

Experimentation like this is how people build fluency and trust in AI. It’s also how they figure out what actually works in their specific context. Our research has shown companies that prioritize adoption and experimentation over designing the 'perfect' AI strategy are 2x more likely to report significant innovation gains.

This is also why we’ve opened ShipIt — our quarterly 24‑hour hackathon where teams prototype and ship new ideas — to teammates across the organization, not just engineering. It’s a dedicated, protected window to tinker with AI on real problems, which is valuable for every function.

That being said, experimentation doesn’t happen in a vacuum. People thrive with the right scaffolding and support. Unfortunately, this is where many organizations are falling short.

Why traditional AI training isn’t working

On paper, it might look like companies are investing heavily in AI capability building. Seven in 10 workplaces say they offer AI training, but employees report that the training is of little practical use.

What actually works is not another 90‑minute learning module, but those small bursts of community, or natural workshops centered on real issues.

People develop skills best when they:

• Bring real problems from their workflow.
• See practical examples from peers and people who work in the same discipline or craft.
• Ask 'silly' questions in a psychologically-safe environment.
• Walk away with something they can apply that day.
• Provide dedicated time to experiment, such as AI days.

This is where HR can fundamentally expand AI adoption from a compliance exercise to a key pillar of how work gets done.

Embedding AI into work and culture

I’ve made it a priority at Atlassian to ensure that AI adoption is a people initiative. We embed AI into strategy, ways of working, and culture, setting company and function-level objectives so every team — from R&D to HR — can see how it fits.

'AI Champions,' as we call them, tailor and reinforce their experiments, share their lessons, and help their teammates adopt AI in practical ways. Our north star — AI is a brainstorm teammate, not a replacement for human ingenuity.

HR should lead adoption and amplify impact

Measuring breadth and depth of AI usage is key. We’ve shifted our focus from simple adoption, i.e. Weekly Active Users (WAU) — to strategic use. We now actively support and measure 'Super Users' — those who complete 40+ actions in a week, to capture intentional use.

HR can amplify impact by incentivizing adoption and elevating champions who experiment, share knowledge, and model meaningful AI use. I firmly believe that these matter more than meeting an arbitrary quota.

Some of our most impactful wins come from HR itself — leading by example. NORA, our onboarding AI agent, reached 93% weekly adoption among new hires, cut onboarding effort significantly, and helped new employees perform at or above tenured levels within three months, all without replacing human connection. There’s also CoCo, an AI agent supporting managers in compensation conversations, which removes operational burden while keeping decisions human, fair, and compliant.

HR and people leaders are uniquely placed to lead AI adoption — and our research backs up the value of doing so. HR and marketing leaders are 2.3x more likely than tech leaders to report meaningful AI-driven business impact.

The takeaway is that HR’s home turf is now the AI era. People leaders must lead adoption, celebrate champions, measure meaningful use, and seize the moment—or the moment will pass them by.