AI belongs in the flow of work. Not in charge of it.
Same technology, opposite outcomes. The difference is not the tool. It is where the judgement sits.
A team leader in an NDIS provider recently told me she had started using AI to write feedback for her team members.
Not to prepare it. To write it. Copy, paste, send.
Her team knew within a fortnight. The language did not sound like her. The feedback referenced things she had never observed. And something quiet broke in the relationship, because the message underneath the message was: you are not worth my actual attention.
Now hold that against a different leader in the same organisation. He uses AI to rehearse. Before a difficult conversation with a support worker about boundary issues with a participant, he practises privately. He tests his opening. He hears how his words might land. He walks into the real conversation prepared, present and entirely himself.
Same technology. Opposite outcomes. The difference is not the tool. It is where the judgement sits.
The line that matters in care and community sectors
In sectors like disability, early learning, aged care and health, the relationship is not adjacent to the work. The relationship is the work. Families, participants and residents can tell when they are being processed rather than seen. So can your staff.
That does not mean AI has no place. It has an enormous place. Your coordinators are drowning in case notes, rosters, reports and compliance documentation. Every hour AI gives back from administration is an hour available for the human work that only humans can do. The organisations getting this right are using AI to protect the relational core of their work, not to replace it.
The organisations getting it wrong are letting AI drift into judgement. Feedback written by a machine. Performance assessments summarised without a leader ever forming their own view. Decisions about people made on outputs nobody questioned.
Three questions before AI touches a leadership task
I give leaders three questions, and they resolve most situations quickly.
Is this preparation or replacement? Using AI to rehearse a conversation, structure your thinking or draft a starting point is preparation. Sending its words as your own is replacement.
Would I be comfortable if this person knew AI was involved? If the answer makes you flinch, you already have your answer.
Does the judgement stay with me? AI can inform a decision about a person. It should never make one.
Why this is a capability issue, not a technology issue
Every organisation I work with is writing an AI policy right now. Very few are building the capability underneath it. A policy tells leaders what is permitted. It does not build the judgement to know, in the moment, on a Tuesday afternoon, which side of the line a task sits on. That judgement is people leadership capability, and it is learnable, like every other capability your leaders were never explicitly taught.
This is also why I integrated an AI coaching companion into my own work through SmartSkills, powered by Onda. Used well, AI gives leaders a private space to rehearse the conversations that matter before having them with a real person. The human stays in the loop. The technology sharpens the leader’s judgement. It never replaces it.
One thing to do this week
Ask your leadership team where AI is currently touching people decisions in your organisation. Not where the policy says it should be. Where it actually is. The gap between those two answers is worth knowing about now, while it is still small.
Stay in the loop
Leadership field notes, straight to your inbox.
New insights, practical tools and free resources. No noise, unsubscribe any time.
Free tool
See where the gaps sit across your leadership layer.
The People Leadership Scorecard takes three minutes.
Take the Scorecard
