AI in HR: 4 Lessons Shared by Global HR Leaders at Outsmart London 2026
The HR teams pulling ahead with AI aren't chasing more dashboards. They're feeding AI the right context, putting it where leaders already work, and spending the time saved on strategy.
"The most important thing about technology is not the technology; it's what we allow it to do." Dave Coplin, Chief Envisioning Officer and author of The Rise of the Humans, opened Outsmart London 2026 with that idea, and it set the tone for the day.
From Standard Life's boardroom wins to GSK's AI roadmap, every session came back to how HR teams are actually working with AI, and how that work is turning people analytics from a function that simply serves insights into one that drives measurable business impact.
We heard from Wendy Cunningham and Olly Britnell of Standard Life, Mads Frank of GSK, Mattijs Mol and Kimmo Paaso of Wärtsilä, Konstantinos Babetas of Titan Group, and Visier's own Steve Holder.
From a day jam-packed with insights, here are the four big ideas I came away with.
1. Insight is worthless until it drives a decision
People Analytics (PA) has spent years perfecting the dashboard, but all too often, the validity of the data becomes the topic of discussion, not the dashboard insights. Worse, many organizations don’t have the capacity to deliver the requested dashboards until long after the decision needed to be made.
Which means leaders are still making the final decisions on best guesses and gut feel. Visier's Vice President of Solution Advisory, Steve Holder, opened his session by stating this problem clearly: "We can continue to generate insights, but if someone generates an insight and nobody looks at it, does anybody care?"
Chief Operating Officer for the People Function at Standard Life, Wendy Cunningham, was even more direct in her session, From Vision to Execution: Delivering Business Strategy Through Insights: "What we know is that trusted data creates value only when it informs decisions, challenges assumptions, and drives outcomes."
Head of People Technology, Digital, Analytics & Change at Standard Life, Olly Britnell, showed what that looks like in practice. Three years ago, Standard Life's People Analytics team measured success by output. It fielded a steady stream of ad hoc requests for dashboards, forecasts, and data science models, with no shared view of which insights mattered.
So the team shifted its focus from delivering data to driving better decisions. They aligned with the leadership team on "the right measures, the right KPIs, the right insights that we want as an organization." Ad hoc reporting has since fallen by more than 30%, freeing the team to do work that shapes decisions instead of just documenting them.
As a final story, Wärtsilä’s Mattijs Mol, VP HR Technology, Strategy & Insights, has proven that PA can earn a seat at the decision-making table when it stops simply serving HR and starts changing business decisions. Mol expects the function to outgrow the name “People Analytics” altogether, evolving into workforce intelligence.
2. AI is only as good as the context you give it
AI is now firmly embedded in business; to some degree, every organization has integrated AI into their workflows. But not every org is seeing ROI, and as I heard at Outsmart, that gap often boils down to missing context.
Gartner draws a direct line between business context and AI performance. It predicts that by 2027, organizations that build context into their AI-ready data could see agentic AI accuracy improve by as much as 80% and costs drop by as much as 60%.
With workforce data, the stakes are higher. Sensitive data dumped into a general LLM without context or governance is a dangerous mix. Mads Frank, Global Head of People Analytics and Insights at GSK, shared his fear of HR business partners exporting raw reports into general-purpose AI tools and asking them questions.
He was far from the only one with that experience, and it's exactly why we built the Workforce Context Engine as the architecture behind Visier's Workforce AI.
The Workforce Context Engine is a purpose-built system that continuously unifies, enriches, and delivers workforce context to any tool, AI agent, or person asking a workforce question. Critically, it ensures every answer shows only what that person or agent is allowed to see, even as reporting lines and org structures shift.
3. AI-Human collaboration: Take the robot out of human roles
AI’s deeper and deeper integration into the workforce was an ongoing conversation at Outsmart. But what I heard most was that AI isn't coming for HR’s job. It's coming for the parts of those jobs nobody wanted in the first place.
Coplin framed this as an opportunity to remove the robotic grind from human work. Towards this end, he has set a new bar for measuring the success of how we use AI across our businesses: "Our success will be measured not by how much time we save, but by what we choose to do with the time that is saved. That is the most important thing that's sitting in front of you.”
For Wärtsilä's Kimmo Paaso, Head of People Analytics, AI-powered automation clears out "the work you don't actually want to do as a PA function anyhow. It's the boring work that you don't really get applause for, and it doesn't really do too much. It doesn't have that much value." AI can create new capacity to give PA room for more strategic projects, like how to save the organization €50M in five years.
Titan Group's Konstantinos Babetas argued that PA teams need to change how their value is measured as AI takes over more routine reporting. The question is no longer how many analysts or hours a report takes, but what kind of impact can one analyst have in one day, in one week, in one month?'
If AI in HR means the robotic work goes away, it only matters if PA transforms the role, (re)focusing on the questions that move the business.
4. AI in HR has to meet people where they already work
If it isn't happening in your business yet, it will be soon. HR's day is moving out of the traditional HR system and into conversations with AI assistants. Coplin put the question to the room: "Why do I need to go and use the HR system? Why can't the HR system be in the conversation I'm already having with my agent about how I'm running my business?"
GSK is already building toward that model. "What I want is a unified front end that is HR consulting and advisory," Frank said. After a period of experimentation, his team has connected an MCP server to GSK’s internal AI assistant, the system employees already use every day.
And it's working. GSK has about 1,500 users on its system today and expects senior leader usage to double or triple over the next two years. But Frank’s long-term goal is bigger still: "I think the real target is to get out to the vast majority of managers where this is the starting point for any strategic talent decision."
Your leaders shouldn't have to leave their workflow to get workforce answers. With Vee, or with Visier's MCP server connected to the AI tools you already use, they don't have to.
The technology is ready. Is HR?
Outsmart London crystallized ideas that have been building in HR since GenAI arrived. AI alone won't transform HR. As global HR leaders made clear, AI needs context and governance, and it has to live where people already work. Most importantly, the time it frees up has to go toward the strategic work that moves the business forward.
In his early session, Coplin reminded us all, “The most important thing about technology is not the technology; it's what we allow it to do, how we work with it, what we enable in our organizations." The HR teams pulling ahead are the ones making that choice now.
Stop guessing and start navigating the AI-driven workforce transformation. Explore the research, tools, and customer stories that help HR leaders turn AI from potential into impact.
