5 Ideas to Redesign Work for a Future Transformation
AI is forcing organizations to rethink how work gets done. These five takeaways from the HR Leaders panel "Redesigning Work for the Human-AI Era" cover where humans must stay in the loop, why ethical guardrails come before rollout, and which metrics actually matter.
This is how Analytics leaders are redesigning work for both humans and AI
If you’re like most HR professionals I work with, the relationship between people and AI technology is top of mind. Which tasks should you and your team handle? Which should you delegate to an LLM? And how do you know when you’re too reliant on one or the other?
This is an important topic. After all, you don’t want your use case to become a headline, like Ford, which reduced its workforce to roll out AI on its assembly lines, only to rehire 300 engineers in a scramble to course-correct.
As Visier’s Principal Researcher, I recently had the opportunity to participate in the HR Leaders panel “Redesigning Work for the Human-AI Era” alongside Arun Muralii, the senior director of people insights and workforce planning at Schneider Electric; Jakob Ejlsted, the head of people analytics at Scandinavian Airlines; and Joost Govers, the VP HR of planning, analytics, and insights at Shell.
Here are five takeaways from these global leaders on what AI workforce adaptation strategies have actually worked.
1. HR shouldn’t redesign work alone
Work redesign is the process of analyzing and improving tasks, workflows, and job descriptions to align with current market conditions, enhance employee engagement and organizational productivity, and meet evolving business goals.
But with AI in the mix, work redesign becomes indispensable. Your organization must totally rethink its processes to determine how employees work and how technology can augment their efforts. HR should play an important role in this process, but it shouldn’t redesign every process on its own.
When the panel was asked how HR leaders should decide which parts of work stay human-led, I said I wasn't convinced HR should make that call alone. Arun Muralii agreed, “When HR does it alone, it feels like policy. When it takes a lead, it feels like a deployment plan.” So, “The people who get most impacted by [work redesign], they need to be part of the co-design.” This is a critical distinction.
To take the lead, as Muralii explained, HR needs to bring together business leaders from all affected departments and facilitate the collaboration. Doing so will allow everyone to deconstruct their work and determine the best ways to add AI into established workflows.
So where does HR add the most value? Two places, in my view. First, bring data and expertise on talent to the table. Second, help set the ethical guardrails. Because some tasks should never go to AI, even when it can do them.
2. Judgment complexity should stay human-led
The human brain is divided into two halves. The left half is logical, analytical, and orderly. The right side is more imaginative and visual.
Jakob Ejlsted keeps this brain chemistry in mind when he delegates tasks to AI or not. During the panel, he asked, “Whether individual tasks can be sequenced in such a manner that we leverage the strengths of AIs and humans.” In real-world scenarios, Ejlsted detailed how he delegates explicit, linear tasks to AI, while performing creative, context-heavy tasks himself.
Joost Govers agreed, “The more work requires ambiguity, ethical judgement, creativity, leadership, accountability—that grey space—the more human involvement remains fairly critical."
Muralii added, “To me, accountability should never live in an algorithm. There should be a named person at the end of it who is accountable when things go wrong. That forces us to think twice before just throwing everything to AI.”
I found myself agreeing with the consensus shared: AI should handle repetitive, linear tasks, while humans should manage imaginative and/or ambiguous tasks that require empathy or ethical judgment.
One thing I also think is worth mentioning: the tasks we're handing to AI are the same tasks early-career employees have always used to build judgment. If AI absorbs them, the apprenticeship path that produced your next generation of senior people goes with it. That's the current challenge talent management teams are facing.
3. Ethical guidelines come first
Every organization should prioritize developing guidelines governing ethical use of AI for its employees, and HR should lead that work, a point I made early and one Murali came back to later, adding a useful test.
Start by listing the tasks employees should never hand to AI: processing proprietary data, delivering personal feedback to a coworker, making final hiring decisions. Add any scenario where a mistake would be too costly (or too human). Then make the guidelines explicit and built to outlast the technology.
In Muralii’s words, "AI is going to keep on evolving, so the thing that's going to hold true for you is going to be your principles. If it's based on things that don't have an expiry date, like accountability or fairness, it's going to serve you well in the long run."
Near the end of the event, Ejlsted added a point worth sitting with: when companies give AI agents names and treat them like colleagues, accountability drops. People pass AI-produced work up to their managers for a second review more often, and they catch fewer errors themselves. The lesson for your guidelines: accountability lives with a person, rather than with the tool.
4. Measure business outcomes, not AI adoption
Many companies measure AI adoption rate, but this is a poor metric. Instead, the consensus among panelists was to measure business outcomes, like greater productivity (output over input) and higher decision quality.
According to Govers, "If AI is just there to make existing work go a little faster, you haven't transformed the work. You've basically digitalized an inefficiency."
I think it’s also worth asking, “What do employees do with the extra time AI gives them?” Considering this metric will inform the way you redesign roles in the future.
5. You can’t plan for what you can’t see, so take stock first
Maybe you have a mandate to automate a set number of activities, save X number of hours, or a list of AI tools waiting to be rolled out. Before any of that, I’d argue for measuring much more basic details: workforce composition, capability distribution, and where skills actually live. From my experience, most organizations still don't have one.
Yet, when you understand the people and skill sets your company already employs, you can make better decisions around roles, responsibilities, and future hiring needs. And, you have a much more accurate idea about where AI can make a real impact on business outcomes, without redesigning work and impacting the humans who do that work, based on mere assumptions.
AI-Driven Workforce Redesign is Built on Three Pillars
Looking back at everything Arun, Jakob, Joost, and I covered, I heard several principles running through the conversation about how organizations can successfully navigate workforce redesign. At Visier, we make this possible with the intelligent infrastructure that fuels workforce transformation by helping leaders:
Understand workforce reality
Adapt to more agile workforce planning
Make smarter, resilient workforce decisions
Transform your work the right way with AI
Every AI workforce adaptation strategy depends on the same thing: an accurate picture of your workforce. But most HR reporting runs on annual cycles and static exports, and by the time the data surfaces, the transformation decisions have already been deployed.
I co-authored Visier's latest research, Hidden in the Headcount: The State of Workforce Transformation During the AI Reckoning, where we analyzed over 3.6 million live employee records across 155+ enterprise organizations to show exactly how AI is reshaping workforce composition: by domain, role, and age cohort.
It's the granular view into how AI is actually reorganizing work that can help inform your own workforce transformation.
