Your job description just expanded: welcome to AI transformation

Across industries, HR leaders are watching their titles change. What used to fit neatly under talent or learning now includes words like transformation, systems, or AI. The reason goes beyond semantics. The scope of the role is stretching to match a new kind of organizational change.

Satu Salminen’s role at ZoomInfo captures this evolution. After years of leading manager and leadership development at Meta and GitLab, she joined ZoomInfo to head global L&D and talent management. Together with her team, she's been an early adopter of AI. In June, Satu stepped into an even more AI-focused role: to lead AI transformation,  a reflection of how people leaders are being asked to architect enterprise-wide change.

We sat down with Satu Salminen for a conversation about what AI transformation looks like in practice.

“Every day looks quite different, which is really enjoyable for me,” she says. “I help our teams, our employees, to get comfortable using AI in a meaningful and safe, responsible way in their work so that they can get more done in their days. Some days, my hours are filled with literally training our teams on how to use AI or finding the most feasible use cases, because not everything should be done with AI, even if it’s possible. Sometimes I’m using more time on change management when I’m driving new initiatives or communications. And one of my favorite things is when I get to build my own AI agents.”

Why HR is leading the next transformation

The promise of AI has outpaced its reality inside organizations. Studies from MIT show that most AI projects fail to meet expectations, often because adoption lags behind ambition. HR is uniquely positioned to close that gap.

At ZoomInfo, Salminen’s team started early. They secured enterprise licenses, built a safe environment for experimentation, and gave people space to learn through use. 

“People often tell me that they’re “not technical" and don’t know where to begin to use AI beyond basic chats, she explains. “So I meet them where they are, share my screen, and walk through the steps and prompts together. Seeing those light-bulb moments changes everything.”

The lesson is that transformation happens at the pace of learning. Ten hours of hands-on time, she adds, is often enough for people to cross from curiosity to confidence.

How habits turn into adoption

Satu brings decades of L&D experience to this new role, which shapes how she handles the emotional side of transformation. “I’m definitely using my good old change management tricks that I’ve picked up over the years,” she says, “like that Kübler-Ross change transition curve, really thinking about the emotions people face at different times of a transition. Some might be feeling a little bit lost and then I try different techniques. Some might actually be really excited about this kind of an opportunity, so then I go in with full gear like, ‘Let me show you how to build an agent.’”

The same principle applies to habit formation. “I think it’s key to use really business-relevant examples,” she explains. “I’m never just walking in and giving them some random examples of how other companies are using AI. I spend time thinking, okay, this is your role, tell me about your key pain points. Some really mundane tasks that take hours of your week. Then I help identify where AI could give you a lot of lift. Relevance is key in building habits.”

She also uses hackathons to turn learning into collective momentum. “It’s helpful to run some hackathons with the whole team,” she says. “Both identifying opportunities one-on-one and brainstorming together. That has helped in building those habits.”

Building guardrails that support speed

Momentum without safety rarely lasts. ZoomInfo has put a few core foundations in place: enterprise-grade tools, clear access controls, and a concise safe-use policy: “a few clear and specific paragraphs are better than fifty pages people won’t read.” “Free tools are free for a reason,” she adds. “The cost of an enterprise license is the cost of keeping your information safe.”

She also designs agents with built-in guardrails. They are told when to answer, when to pause, and when to admit uncertainty. It’s the digital version of management: clear expectations, context, and boundaries.

When HR thinks like product

Satu’s approach borrows directly from product teams. Every potential use case is plotted on a two-by-two matrix: business impact on one axis, difficulty on the other. High-impact, low-effort ideas rise to the top; complex ones move later in the queue.

The same product logic guides build-versus-buy decisions. “If you have engineers and AI expertise inside your company, building may create leverage,” she says. “If you don’t, buying gets you speed. Either way, talk to customers first and look for truly AI-native vendors.”

This prioritization discipline turns ambition into an executable roadmap, helping HR teams show progress while learning what works.

Knowledge management as a quiet advantage

The discussion around AI often skips a critical point: the quality of internal information. “If you put in outdated data, outdated documents, and expect the AI to come up with updated results, no, it won’t do that,” Satu says. “I make sure that whatever I upload when I’m building an agent, I glance through and make sure it’s up to date. If something isn’t, I update it immediately or ask someone else to. I don’t put anything in if I know it’s not current.”

Her team updates information in phases, only what connects to active projects. “Fix what you touch, then move on,” she adds. The principle is pragmatic: don’t wait for a perfect data set before you start building capability.

She treats AI systems like smart interns: capable, quick, but dependent on clear guidance and good input. It’s a useful metaphor for how leaders can blend trust and oversight.

The coalition driving transformation

When asked what makes this new era of HR transformation work, Satu’s answer is pragmatic: collaboration. “It’s not just IT and HR that need to be playing along,” she says. “It’s also engineering, it’s also legal, because there are so many privacy concerns that need to be alleviated. We’re all working closely together because AI is really shifting how work gets done.”

This model marks a turning point for HR. The department once viewed as a people partner is becoming an organizational architect, responsible for how humans and technology build capability together.

A shift that starts with people

Satu’s closing advice captures the balance of urgency and care every HR leader faces. “Start from small,” she says. “Analysis paralysis might be preventing a lot of folks from doing more productive, advanced things with AI. Find a way that works for you and your company in a safe, responsible way. Meet people where they are, because folks will be at different levels of their AI journey.”

She also reminds peers to lean on each other. “There are so many of us who are figuring this out. More than ever in my career, I’ve noticed that folks are really willing to share their best practices. Don’t try to figure it out alone.”

The HR job description has expanded, but the through line remains the same: enable people to thrive through change. The difference is that the change now includes systems, data, and an entirely new kind of teammate.