A few months ago I sat down at my desk and tried to understand what AI really is.
Not the headlines. Not the breathless predictions about robots taking everyone's jobs or the equally breathless reassurances that nothing would change. The actual thing. What it does, how it does it, what it gets wrong, and — most importantly, because I am a lawyer and this is apparently how my brain is wired — what happens when it goes wrong to someone who was not expecting it to.
I had been away from practice for a while. A health issue, now resolved. The kind of absence that turns out to be useful in retrospect, because when you come back to something after a gap you see it with different eyes. You notice what changed while you were not looking. And what had changed, more than I expected, was this: AI had moved from a theoretical future to a present-tense reality, and most of the legal and employment world had not caught up to it yet.
That observation became a question. Then the question became a business.
The specific thing that bothered me
I am not, by nature, a technologist. I have no particular gift for code or systems or the mechanics of how software works. What I do have, after more than twenty years in employment law across South Africa and New Zealand, is a finely tuned instinct for what happens to people when institutions make decisions carelessly.
Employment law is, at its core, about power. About the relationship between employers and the people who work for them, and about what happens when that relationship breaks down. I have spent most of my career in the space where things have already gone wrong — investigations, grievances, mediations, the procedurally defective dismissal that lands on a Friday afternoon. I know what carelessness costs. Not in the abstract. In specific, documented, human terms.
So when I started reading about how AI tools were being used, in hiring, in performance management, in workforce restructuring, the lawyer in me immediately started asking the questions nobody seemed to be asking loudly enough. Who is accountable when the algorithm gets it wrong? What does good faith consultation look like when the decision was made by a model? What does a New Zealand employer actually owe their staff before they deploy a monitoring tool that watches what they do all day?
These are not hypothetical questions. They are the questions that will be sitting in Employment Relations Authority waiting rooms within the next few years. Some of them already are.
What I am not
I want to be clear about something, because I think it matters for what this space is going to be.
I am not here to sell AI. I am not an evangelist, a disruptor, or a person who uses the word "ecosystem" without irony. I am genuinely excited about what these tools can do — I have seen enough now to know that the practical possibilities for legal practice are real and significant — but excitement without judgment is just noise. The legal profession has enough of that already.
What I am trying to do is something more specific: to bring a practitioner's eye to a conversation that has, until now, been dominated by technologists on one side and anxious sceptics on the other. Neither group is wrong, exactly. But neither is asking the questions that matter most to the people whose working lives are actually affected by these decisions.
Safe and ethical AI use is not a compliance checkbox. It is the difference between a tool that serves people and one that quietly disadvantages them while everyone assumes the algorithm knows best.
That is why I am doing this. That is what Lex Praxis is for.
The invitation
I am new enough to AI to remember what it felt like not to understand it. I think that is an asset, actually. It means I can explain it without condescension, and I can admit uncertainty without it feeling like a confession.
What I am building here is a conversation. About AI, about employment law, about the intersection of the two, and about what it means to use powerful tools responsibly in a profession built on trust.
And I'm inviting you to come along for the ride. It is going to be interesting.