Everyone keeps saying that AI — ChatGPT, Claude and similar tools — is saving them time, making them more efficient, and freeing up their working hours. But what if I told you that I now spend 50%, or even more, of my time reviewing and rewriting client texts that were, of course, generated by ChatGPT — and checking writers who submit AI-created content as well?
I’ve spent years in PR, enough to remember when the work still had a bit of blood in it. The best part for me was always the chase — finding the angle, the hook, the satisfaction of landing a story in exactly the right outlet, because this industry used to run on sharp thinking.
Now a good part of my week is spent opening documents that are, on the surface, perfectly respectable — fine structure, clean grammar. But the whole thing has that college-report sheen so beloved by AI and so often mistaken for competence. There’s no pulse, no real point of view behind the words.
Using AI Is Not the Same as Being Competent
Let me be clear about where I stand. AI is going to transform this industry, and I want it to. I disagree with the experts who insist a machine can never do what a PR professional does — that’s a comforting story, and comforting stories are how you get overtaken. Any business treating AI as a passing trend will fall behind.
But adoption is not the same as competence. And right now, almost everyone in PR is using this tool badly.
I run a PR agency. I didn’t build it to spend half my week quality-checking a chatbot.
That’s not AI’s failure. That’s ours. We pointed the most powerful tool we’ve ever had at the one part of the job it’s worst at.
Content Quality Drives PR Performance
Yes, the data backs up AI’s usefulness. Microsoft and others have shown it performs well across standard formats — press releases, pitches, social content. But who is actually checking the quality? Just because a press release meets the minimum bar to be published doesn’t mean it’s worth reading, that it has impact, or that it serves the audience.
In my experience, you don’t win attention by being correct. You win it by being interesting and a little unexpected. When a founder bypasses the chatbot and simply talks to me, I get more in twenty minutes than from a month of polished drafts — the difference is visible the moment a real human being speaks.
What AI is excellent at is producing the average of the internet, and very bad at producing anything with appetite.
Without Communication, AI Content Falls Flat
I keep a folder with hundreds of screenshots where I show clients that their content is 50–70% AI-produced — not even 10%. I do it because we share the same goal: make the PR effective. I’m open with them about it, even knowing some may see me as rude. But when I discuss the same content in person — with professionals, colleagues, founders — it becomes far more interesting and worth the time. That’s why I put communication first in a world of accelerating AI: it’s the best way to learn more than an AI-generated thought.
Moreover, as reported by MIT Technology Review and The Verge, large language models work by recognising and recombining patterns from the information already in their training data. In other words, they’re built to reproduce what exists — not to hand you a point of view nobody has had before.
Right now, many trends are being shaped by AI tools that lean on existing information rather than generating genuinely new thinking. “Everything new is just a well-forgotten old idea,” as the saying goes — but the human mind is far smarter and more creative when it comes to coming up with original insights, connecting ideas in unexpected ways, and pushing thinking beyond what already exists.
When I worked in television writing scripts, we also had to come up with ideas that had never been used before. Of course we used Google — but only as raw material, as facts. We would take that information, go deeper, look into the details and nuances, and then build something new from it. Not just combine and summarise the internet. That process is what creates original thinking. And that’s exactly what I feel is missing now — a truly new point of view.
Media companies aren’t the only ones letting journalists go because of AI — agencies are cutting writers and analysts too. Why? Because everyone is using AI for content, so one manager can now do the job of a whole writing team. I believe many media companies will eventually bring experienced journalists back, once they realise that AI can generate content but can’t replace a distinctive writing style, strong editorial judgment, or an authentic voice.
AI Is Progress. Make Sure It Stays That Way
If you ask me what’s actually in short supply now, it isn’t writers — it’s people who can think critically about what they’re producing. People with accountability and a voice of their own, who can look at AI-generated content and ask: does this make sense, does it say anything, does it achieve what we need?
If we outsource our thinking, our professionalism and our lived experience to a machine, we risk losing the very thing that makes progress possible — original human insight. And if we don’t truly adapt, but just blindly use AI, then maybe we should start taking Elon Musk seriously when he says digital intelligence will surpass us all by the early 2030s.
Because for all the automation and efficiency, the best ideas still come from people. From conversations that go off-script. From debates that get messy. From moments you can’t predict — when someone says something that makes you stop and reconsider everything you thought you understood.
If we want to build meaningful communication, strong PR and viral trends, we have to keep human thinking at the centre of it. That’s where the work is. Everything else is just formatting.
