There’s no questioning that AI has thrown a big old stick of dynamite right in the midst of the way work is created. For editorial teams, the subsequent content explosion has brought with it a lot of questions and, if we’re honest, a hearty side helping of existential angst.
It’s transformed workflows irrevocably, which isn’t surprising in industries where time is a currency and AI platforms can churn out thousands of words quicker than you can open a thesaurus.
But the thing about the word ‘transformation’ is that it’s neither inherently positive nor negative. At its core, it means a dynamic, fundamental change—which is definitely true. But with change comes—and this is a key word—nuance. Ironically, that’s something AI-generated copy tends to remove.
So here are some thoughts on the AI copy debate. Except instead of talking in statements, we’ve done our best to inject some nuance back into things. Because like most things in life, there are good bits and there are less-good bits. The key is staying informed and using your own (human) judgement to use the better bits to your project’s advantage.
AI can accelerate editorial workflows, but not always
It’s true that AI often speeds things up. (Yes, we’re aware that’s a statement, but hey, there’s that nuance for you.) In doing so, it’s also helped lower barriers to creating written content. But it doesn’t necessarily follow that it’s lowered the barriers to creating good written content.
That’s partly because good writing has never been about words alone. For a piece to truly speak to its reader, the experience has to feel seamless. There’s a compelling flow that takes you from one thought to the next in a way that feels intentional. And the way those thoughts are expressed—voice, tone and language—feels clear and authentic. You finish having digested a cohesive point that’s been delivered in an interesting way, whether you agree with what’s been posited or not.
Now, AI is undoubtedly helpful for some of that stuff. When you’re struck with a case of writer’s block, or are struggling to unpick a complex idea, it’s great for giving you a framework for your thinking and streamlining that brainstorming process.
Where it falls down, however, is when you ask it to generate the piece from scratch. More often than not, you’ll find the finished product circular in argument, peppered with dubious-sounding claims and written in a way that feels weirdly familiar. As the ultimate people pleaser, it tells you what you want to hear, not the objective truths that make your writing better.
As a result, it’s not uncommon to find yourself spending just as much time editing a piece to make it feel ‘human’—and fact-checking that those claims are actually legit—than if you’d just written it yourself in the first place. Which brings us on to our next point.
Navigating the sea of sameness
Noticed how everything’s starting to sound weirdly familiar? Us too. We’ve all got stuck into the em dash debate and found ourselves noticing the repetitive use of certain buzzwords and sentence structures. (We’re talking about you, rule of three; and you, ‘It’s not X, it’s Y.’ We’ve also never been asked to ‘dive’ into anything quite so much… call us Tom Daley.)
Jokes aside, it’s no coincidence that there’s been a marked homogenisation of the tone and style of vast swathes of the comms we consume in recent years. It’s also a bit of a Catch-22: as more AI-generated content gets published, the more we ingest—and the more we subconsciously begin to emulate its style and turns of phrase. It’s not just that we’re using more AI, it’s that we’re starting to imitate the imitator. See what we did there.
That’s why one thing in particular has become more important than ever: sounding like an actual human being with a tangible personality who knows the people they’re speaking to.
Authenticity as the gold standard
What hasn’t changed is that people generally still don’t like to be marketed to. People are also developing a distinct aversion to copy that’s obviously AI-generated—something often referred to as ‘AI slop fatigue’.
We see it as a relationship thing. Like all good partnerships, the bond between brand and customer is built on trust. So why should your customers trust what you’re saying if it’s clear you haven’t taken the time to make sure it’s something that actually speaks to them?
As humans, we have something AI doesn’t, and that’s direct experience of living in the real world. As marketers, those human, real-world conversations with our customers translate to insights—language, humour, nuance—that allow us to resonate with the specificpeople we’re talking to. AI struggles to do that.
When used at its most effective, AI is an eager-to-please assistant. And there’s no shame in letting it help you, even if it makes a rubbish cup of tea. But when you let it take over completely, the quality of your message suffers because the personality gets stripped right out. That’s a surefire way to make sure your message blends in with every other that’s been generated in the same way.
Looking to the future
Change is the only constant, and AI isn’t going anywhere. Not that we’d necessarily want it to; when used in the right way, it’s genuinely helpful.
But for an editorial project to be successful, its team need to know how to work with it. It’s a partnership, but not an equally weighted one—the fact remains that no algorithm can fully replace the intricacies and peculiarities of the human mind.
Ultimately, publishing AI-generated content without proper editorial oversight is like trusting a toddler to decorate your house. Unlikely to end well, and you’ll be the one left staring at the consequences.
And there’s an analogy no AI platform could come up with.
Related thoughts
