At a glance
- AI adoption at scale is a cultural and organizational challenge, not a technology rollout
- Sustainable AI transformation starts with shared ownership, not centralized control
- Moving from AI experimentation to real adoption requires clear processes and accountability
- Measuring AI’s impact in HR is possible, but it can often be messy and incremental
- Responsible AI adoption relies on mindset, structure, and people-first leadership
Why AI adoption is the real challenge
Access to AI tools is no longer the primary barrier for most organizations.
What’s harder is turning early experimentation into something teams actually rely on.
Across HR and People Ops, the AI conversation has shifted. Most organizations are no longer asking if they should explore AI. They’re asking how to move forward responsibly, without chasing hype, overwhelming teams, or rolling out tools no one actually uses.
And what does all of this add up to? Real pressure to “do something with AI”.
Yet many organizations stall after their early bouts of excitement. The AI experimentation phase produces all sorts of pilots and demos, but adoption rarely follows. AI readiness in organizations turns out to be less about access to technology and more about whether people know how to use it in their day-to-day work.
Episode 2 of AI Mind Talks offers a candid look at this reality.
Rather than focusing on tools, it explores what it actually takes to build an AI adoption strategy that scales—and one that’s grounded in culture, human-first ownership, and proper process.
From “AI projects” to an AI mindset
One of the earliest lessons in HiBob’s AI transformation strategy was that AI couldn’t live as a series of isolated projects.
Early conversations weren’t about features or platforms. They centered on a broader question: “How do you build an AI mindset that lasts?”
The conclusion was clear. Treating AI as a one‑off initiative—a single project, an innovation sprint, or a dedicated “AI day”—wouldn’t create lasting change. AI needed to become an ongoing capability.
That realization led to the creation of a defined AI process, driven by four motivations:
- Building a continuous AI mindset, not episodic experimentation
- Creating guardrails for safety, ethics, and responsible AI adoption
- Maintaining focus and prioritization in a fast‑moving space
- Democratizing AI beyond a centralized tech team
But, as Michal Lewy Harush, HiBob’s CIO, put it during the episode: “It’s very easy to get lost on this journey. So the process becomes the compass, helping us make sure we are focusing on the right things.”
For leaders shaping an AI transformation strategy for HR, this shift is foundational.
At its core, AI adoption goes beyond just a simple case of technology selection. What we’re seeing is that it’s much more successful when it starts with shared understanding and intent.
Democratizing AI: Giving ownership to the people closest to the work
A defining element of our approach was redefining ownership.
Rather than framing AI through heavy governance, we intentionally focused on process. Governance felt restrictive, whereas process invited participation.
As Alon Arbiv, Director of AI at HiBob, put it during the episode: “We decided early on not to use the word governance. It’s a negative word. The moment you start with governance, people stop playing.”
Every team member could propose an AI idea. Anyone could become an “Agent Owner”—responsible for identifying a real pain point, experimenting with a solution, and shaping how it fits into everyday workflows.
This wasn’t about decentralizing control for its own sake. It meant we could place ownership with the people closest to the work, as they’re the ones who know it best.
Team members understand where time is lost, where friction exists, and where AI productivity in HR—and beyond—can actually improve outcomes. When AI use cases for HR originate from those lived experiences, adoption becomes far more likely.
AI moves people beyond task execution and into problem ownership—with the autonomy to shape, refine, and improve how the work actually gets done.
For HR leaders, this offers a practical answer on how to drive AI adoption among employees. And that’s to involve them early, give them ownership, and make freedom and experimentation a central part of the role.
The AI adoption roadmap: From idea to impact
Democratization alone isn’t enough—AI adoption at scale requires structure.
At HiBob, this became a five-stage AI adoption roadmap.
It wasn’t built as a strict framework. You can think of it more as a way of working—something teams could come back to and help move an early idea along until it’s been tried, shaped, and is actually useful in day-to-day work.
- Stage 1. Ideation and experimentation. Anyone can submit an idea. Speed matters, so the goal is to test quickly, learn fast, and be at peace with the fact that many ideas won’t progress.
- Stage 2. The decision point. This is where experimentation becomes more of a business decision. Leaders evaluate which ideas should move forward based on workflow fit and potential impact. Here, the ownership clearly sits with the business.
- Stage 3. Build. Some solutions move through a fast track—no‑code or low‑code assistants designed to deliver immediate value. Others need a bit more programmatic, scalable approaches. In any case, the leaders stay deeply involved, providing context and defining what success looks like.
- Stage 4. Adoption and enablement. We know that AI only delivers value when it becomes part of daily work. That’s why key areas like training, visibility, and enablement are treated as absolutely essential.
- Stage 5. Ongoing performance review. Every agent is revisited regularly. Is it being used? Does it still meet our expectations? Does it reflect how the business operates today?
For HR leaders, the takeaway isn’t necessarily to replicate this roadmap exactly. It’s to recognize that moving beyond experimentation requires clear ownership at every stage.
Recommended For Further Reading
Adoption is the hard part—and measurement is harder
Across the episode, one message came up repeatedly: adoption is the hardest part of AI transformation, as even well‑built AI solutions fail if they sit outside real workflows.
To address this, at HiBob, we introduced a central agent directory—a shared place where people can discover available AI, understand its purpose, and learn how to use it.
Each agent has a named owner, clear goals, and visible usage data. And recognition plays a role, too. Sharing and celebrating agents and the value they bring reinforces the very real idea that AI is a collective effort.
Measuring AI impact in HR, however, remains complex.
Some use cases allow for direct measurement—time saved, volume handled, coverage increased. Others create value that’s harder to quantify, such as better decisions, faster learning, or new capabilities that didn’t previously exist.
In practice, this means not every AI use case maps cleanly to a dollar value or a single KPI—and that’s fine. Some of that value shows up elsewhere. Better decisions. Faster learning. Or entirely new capabilities that simply weren’t possible before.
For HR teams, this perspective is critical.
Measuring AI productivity in HR is much more about understanding whether people are genuinely working differently—and better—over time.
What HR leaders can take away
HiBob’s AI Mind journey doesn’t offer a single blueprint—and that’s intentional. But it does surface practical lessons HR and People Ops leaders can apply:
- Start with real pain points, not novelty
- Treat AI as a cultural shift rather than a traditional tech rollout
- Assign ownership early and clearly
- Normalize experimentation and failure as a big part of the learning process
- Measure what matters, even when it’s imperfect
At the end of the day, AI transformation best practices are about intention—not speed.
If you’re refining your own AI adoption strategy and navigating uncertainty, false starts, or measurement challenges, then trust me—you’re not alone.
Join me and my guests, Alon Arbiv, Director of AI at HiBob, and Michal Lewy Harush, our CIO, for the full Episode 2 of AI Mind Talks. We go deeper into the conversations behind these lessons, and discuss the realities that shaped them.
Because sometimes, responsible AI adoption starts with asking better questions, not chasing faster answers.
