Editorial
When we’re unsure what a new tool can do, we start by giving it what we’re tired of doing — repetitive tasks, inbox clutter and everyday frustrations that feel like a waste of time. In short, the drudgery. Especially when the tool promises speed beyond anything human.
But that instinct can send us in the wrong direction. When considering where to implement AI in HR systems, we need to start with what we are trying to accomplish with our systems, not just what tasks are annoying.
Evaluate Before You Automate
Every task lives inside a system: a network of processes, policies, norms, assumptions and signals.
So before you automate a piece of work, you have to ask, “What is this doing inside the system? What is it compensating for? What is it signaling? What breaks if it disappears?”
If we apply AI to our systems reflexively, if we automate based on effort instead of intent, we risk reinforcing poorly designed systems.
For example:
- Automating a poor performance review process won’t make it more fair.
- Digitizing a convoluted onboarding experience won’t make it more welcoming or less confusing.
- Speeding up candidate screening won’t fix bias if your criteria are flawed to begin with.
When you start by prioritizing AI adoption towards the drudgery, you’re optimizing the past. When you start with your purpose as a function, when you look at the system as a whole, you open space for reinvention.
Instead of asking “what’s repetitive?” HR should ask:
- What kind of work are we trying to protect, and what kind of work are we willing to rethink?
- What work is core to our mission vs what has just been inherited from legacy systems?
- What work is legally required to meet regulatory obligations?
- Where does human judgment add unique value and where does it introduce bias, delay or error?
- What processes are misaligned with the employee experience we say we say we want to deliver?
- Where are people wasting time on translation between systems, expectations or policy?
These aren’t hypothetical questions, they are strategic filters. They help HR teams move from automation-as-efficiency to automation that reinforces alignment both with organizational goals and human needs.
A Framework to Help You Decide
To use AI intentionally, we should think like system designers, not system operators. Our job as human designers is to check if the process we are automating is worth keeping in the first place. A framework that has been circulating for decades in process engineering circles can help us here.
Eliminate —> Simplify —> Automate —> ... with a fourth step, Augment, to reflect the uniquely collaborative nature of modern AI systems.
This simple framework protects process integrity while determining where speed can be increased. Automation is at its heart a design decision. Bad design, once automated, becomes significantly harder to notice, to challenge or to undo. This framework isn’t a checklist, but a pattern of thought.
Here are the four pieces of the framework:
- Eliminate what no longer serves a purpose: Is this task necessary at all? What assumptions is it based on and are those assumptions still true?
- Simplify what’s worth keeping but doesn’t need to be complex: Could this be redesigned to reduce complexity or confusion before we introduce tech?
- Automate what can be offloaded without eroding context or care: Can AI support or take over this task without losing nuance?
- Augment the work that depends on human judgment, nuance or trust: Would a copilot model (where AI supports, not replaces, the human) lead to better decisions or deeper insights?
Know What to Scale (and What Not To)
The framework gives us a way to slow down, to orient our plans around meaning and mission and to make sure what we scale is worthy of being scaled.
Looking at your systems through this framework will help you avoid scaling bad systems. It will also surface work that feels like busy work, but is actually an early warning sign of process debt. You will probably find places where AI won’t be helpful. And some work won’t fit neatly into any of the four categories because it needs to be redesigned from the ground up.
When we talk about “the system” we’re not referring to your tech stack or a specific platform. We are talking about the full ecosystem that shapes how work gets done.
Conference
Sep
16
HR Tech Conference Las Vegas 2025
Register
Conference
Oct
27
Gartner HR Symposium/Xpo Orlando 2025
Register
Processes that govern decisions and workflows, policies that dictate what’s allowed or assumed, language that frames what matters, the power dynamics that determine who gets heard — all are parts of “the system.”
The Real Opportunity of AI in HR
The real opportunity of AI in HR isn’t speed, it’s system clarity. Done right, thoughtfully identifying targets for AI implementation can not only make our work better but can also help us see where work is misaligned, decisions are delayed or energy is being misspent. But that will only happen if instead of asking “what’s tedious?” we ask “what’s essential?”
Editor's Note: Read more about AI in HR technology and processes below:
- If We Want AI to Help HR, HR Has to Join the Conversation — Engineers are designing AI systems to address problems that are rooted in the very systems HR understands best.
- Rethink Your HR Strategy to Include AI Agents — As AI agents make their way into our workforce, HR is called to adapt its workforce strategy.
- Why HR and IT Must Join Forces for AI to Succeed — The overlooked partnership at the center of real AI adoption.
Learn how you can join our contributor community.
