The problem with AI-written HR

AI is making it much easier to produce HR content, but I’m not convinced that that means the content itself is getting any better.

By HR content, I mean the sort of things many of us now use AI to help with every day: job adverts, employee communications, policies and guidance, training material, survey commentary, reports, presentations and, of course, articles and social media posts.

A lot of this material is perfectly competent. It is clear, well-structured and grammatically correct, but increasingly it also has a familiar feel to it. The same sort of phrasing crops up, the same sentence patterns, the same slightly over-confident tone, and after a while quite a lot of it starts to sound as though it could have come from almost anywhere.

AI is genuinely useful. I use it myself to get ideas moving, to challenge the way I have expressed something, to help with structure or simply to turn a rough collection of thoughts into something more coherent. The danger comes when we accept that first polished version as the finished article.

In HR especially, what we write is rarely just information. It reflects judgement, experience and a view about people, our organisations and the realities of working life, but it should also say something about the culture, character and personality of the organisations we work in. If all of that gets smoothed away, the writing may still be technically good, but it becomes much less distinctive and, in my view, much less valuable.

For me, the best use of AI is not to hand over the thinking, but to use it to help make your own thinking clearer and better expressed.

A simple test is this: if your colleagues read something you had written without your name on it, would they still know it was yours? If not, some of your voice has probably been lost along the way, and with it some of the character and distinctiveness of the organisation you are writing for.

What are your thoughts on this?