Job alert: AI is already doing some of HR's work but could it do it all?

Walk into most companies today and you might find AI everywhere in people management. It’s screening resumes, analysing how candidates speak in video interviews, tracking who’s engaged and who might quit. The sales pitch is always the same: faster, smarter and more objective. And honestly, a lot of it works pretty well.

But I keep having this thought experiment: what if we took it all the way? What if your entire employee experience – from application to exit interview – never involved talking to an actual human being? You would interact with chatbots, get feedback from algorithms, have your performance measured by software that tracks everything from your keystroke patterns to how often you check email after hours.

It sounds efficient. Maybe even fair, since algorithms don’t play favourites or have bad days. But it also feels like we would be missing something, though I’m not entirely sure what.

I’ve started calling this the HR margin – that space between what’s measurable and what actually matters when you’re managing people. Take my friend whose company flagged her as a potential flight risk because she was taking longer lunch breaks. The algorithm wasn’t wrong to notice the pattern. But it couldn’t know she was spending that time helping her father through chemotherapy. A five-minute conversation would have cleared that up.

This gets me thinking about what I’m calling the post-human workplace. Not that robots are taking over, but that more of our work experiences are getting filtered through systems that think very differently than we do. They’re incredibly good at spotting trends and making predictions. They’re just not great at understanding why humans do the things we do.

Which raises an interesting question: who benefits when workplace decisions get handed over to algorithms?

There are legitimate advantages here. AI can catch biases that humans miss. It doesn’t get tired or play politics. It can process way more information than any HR person could. I’ve seen recruitment algorithms that genuinely helped companies hire more diverse candidates by focusing on skills rather than background signals.

But there is a flip side.

A friend runs talent acquisition at a tech company, and she told me their AI system consistently rates candidates lower if they have gaps in their employment history. The algorithm learnt this pattern from past hiring data, but it doesn’t understand that sometimes people take time off for family, education, or just figuring out what they want to do next. Human stuff, basically.

Maybe the real question isn’t whether AI should be involved in HR – that ship has sailed. It’s how we want to blend human judgment with machine intelligence. Do we use AI to replace conversations, or to make conversations better? Do we automate decisions, or do we automate the boring parts so humans can focus on the decisions that need human insight?

I don't think there is a simple answer, partly because different people want different things from work. Some employees might prefer the consistency of algorithmic management. Others need that human connection to feel valued and understood. Generational differences probably play a role too.

But here’s what I keep coming back to: the best workplaces I know are the ones where people feel seen as people, not just as contributors to some larger optimisation function. They’re places where someone notices when you’re struggling, celebrates when you grow and treats your career as more than just a series of performance metrics.

Can AI help create that kind of environment? Maybe. Can it replace it? I have my doubts.

As we figure this out, I think we need to hold on to a few things. Keep humans involved when decisions really matter to people’s lives and careers. Use technology to get better insights, not to avoid having conversations. And stay curious about what we might be losing along the way.

The future of work is going to involve both people and algorithms – that much seems certain. But we still get to decide what that partnership looks like. Do we want workplaces that are perfectly optimised, or do we want workplaces that are perfectly human?

Maybe the answer is that we need both, but we haven’t figured out the balance yet. That’s a conversation worth having.