When robots write the headlines: why storytelling still needs a human touch

A robot holding a pencil and writing in a notebook, next to a fake newspaper with the headline 'The sky is falling!' Overlaid text says 'When robots write the headlines'

The robots are writing the news.

The BBC, the Mail, and a smattering of other outlets are experimenting with AI-generated articles, feeding machines press releases and data to churn out “news” at lightning speed.

No need for coffee breaks, no complaints about the rota, and no union rep muttering about overtime.

On paper, it sounds efficient. In practice? The results are often… let’s say mechanical.

Clunky phrasing, comedy gold

AI-written articles have already served up some unintentionally funny moments.

We have seen weather reports that promised “a chance of wetness”, sports roundups describing results as “one team got more goals than the other” and articles mistakenly listing food banks as tourist attractions.

Then there’s the more serious blunders.

Business Insider recently retracted dozens of essays from supposed freelancers like Margaux Blanchard and Nate Giovanni.

Why? Internal investigations revealed they likely never existed – and the articles contained bizarre inconsistencies and possibly AI fabrication.

Then there was Italian paper Il Foglio which ran an entire edition written by AI – without initially telling readers. The result? Plagiarism from The Atlantic, invented facts, and text so bland that editors needed to step in to clean it up before going live.

Meanwhile, an AI-powered chatbot at Air Canada mistakenly told a customer they could retroactively claim a bereavement fare – something the airline doesn’t actually offer. The tribunal ruled in the customer’s favour, ending in both embarrassment and financial loss for the airline.

And let’s not forget the AI-generated piece on MSN which described recently deceased NBA player Brandon Hunter, as being “useless at 42,” triggering a social media backlash that no marketing team could have planned for.

Even ChatGPT has made errors.

Despite being billed as “PhD-level,” OpenAI’s GPT-5 stumbled spectacularly on basic tasks – miscounting the letters in “blueberry,” inventing fictional U.S. states like “New Jefst,” and bungling spelling of the “Northern Territory.”

ChatGPT has also been accused of making up entire papers, convincing but false quotes, and even correct-sounding but entirely fictional scientific theories. In one case, a law firm cited ChatGPT-generated “legal cases” – that didn’t exist. Imagine the courtroom reaction when the judge asked for proof and got crickets instead. Cue potential defamation suits.

Panic in the newsroom

Predictably, AI experiments have sparked a wave of “robots are taking our jobs!” panic. Journalists everywhere are bristling, and PR professionals are bracing themselves for a future where machines could chew through press releases faster than a graduate on their first Red Bull.

But let’s be real: robots don’t yet know when to push a line for humour, when to hold a beat for drama, or when to spot the story behind the story.

And nuance aside, let’s not forget that AI doesn’t really understand reputational risk. Imagine a brand crisis being left to a bot:

AI headline: “Company X denies poisoning customers, insists only minor sickness.”

Translation: congratulations, you’ve just escalated your crisis from a local complaint to a national outrage.

Tone matters. Empathy matters. Knowing when to apologise, when to reassure, and when to step away from the keyboard matters.

Robots are great… but not great storytellers

Of course, none of this is to say AI has no place. It’s a brilliant assistant. It can crunch data, generate drafts, and help brainstorm ideas.

And at Satsuma, we’re not afraid of the tech – we use it ourselves. But we use it like you’d use a food processor: handy, fast, great for the basics.

Would you serve guests a raw chopped onion and call it dinner? No. You season it, cook it, plate it. The same goes for storytelling.

The human touch will always win

The future of PR isn’t about humans versus robots. It’s about humans using robots wisely, then adding the flair, judgement, humour, and empathy that make stories land.

Because when you strip it back, people don’t remember the clunky, machine-made sentences. They remember the headlines that made them laugh, cry, or think.

And until robots can master irony, satire, or even basic accuracy… we’ll keep our jobs, thanks very much.

Get in Touch with Us

Discover how our creative solutions can elevate your brand and achieve your marketing goals.