Why hyper-personalisation could undermine the authority publishers rely on

In attempting to give people what they want, could we lose the curatorial function of news gathering?

Listening recently to an old interview with The Lord of the Rings author J.R.R Tolkien, he offered an observation about the uniqueness of individuals that explained the difficulty of writing for anyone but himself. “Every single person in the world has a different spoken voice you know, we’re all unique.”

This observation was punctuated by the pleasing sound of puffing on a pipe, which added a considered air often missing in our more fast-spoken times. Tolkien was a philologist first, a lover of languages, as anyone who has read his works can understand, but yet his broad learning could come down to seeing what the individual meant, and how each of those unique human voices contributed to the whole.

Allow me a jump here, as reading through some ideas about audience capture, this tangent of thinking about individuality led me to considering what we mean by “personalisation” in publishing and whether by attempting to do so, we are trying to personalise against an average person and therefore personalising against no-one.

The primary issue with personalisation is actually one of definition. We go from “you read about fly fishing so here’s some extra fly fishing stuff” to “here’s a whole homepage curated to your general interest pattern” and all points in between. It means different things to different publishers, and bearing in mind the multiplicity of audiences, it should. In broad terms, we’re trying to give the people what the people want. It’s a technical challenge Glide CMS has risen to on many occasions, with our unrivalled content categorisation capabilities (shameless plug, well done Rob. Ed.).

A new and deeper form of potential personalisation is now a prospect, “hyper-personalisation” and yes dear reader, GenAI features as the enabler. One senior digital news executive has recently raised this idea through the prospect of having different versions of the same story available to different readers based on their known preferences. You can see how this might work technically, although what it would contribute editorially is a tougher question.

With something like Sports reporting, one can see a use. If for example, you’re interested in the performance of a particular up-and-coming athlete on the basis they’re from your country, then the fact they came third in the event and all the coverage is about the winner, you might prefer a story that goes in on what a prospect they are. Could that be extracted from the canonical write and then rewritten using an AI tool to reflect your preferences? It’s possible, although how clumsy it might be and how it would affect the flow and quality of the existing write is open to question.

It gets much harder when it comes to hard stuff though. Conflicts for example. History isn’t necessarily written by the winners - ask the Mongols - but conflicts inevitably have the stories of at least two sides as part of what has made them a conflict in the first place. Questions of ethics and bias already revolve around such reporting, without adding a machine-enabled bias layer, and you can bet your bottom dollar that bias is already trained in, just ask the recently and quite publicly re-educated Grok. For many publications who set their commercial stall out on the basis of authority, having such variability drawn from a single content source eats away at the very foundations of what they actually are. In short, you pay them to tell you what’s important, not to pander to your biases.

With that said, it’s possible to offer another use of “personalisation” as it pertains to our industry and it’s one we already know all about. In fact, I’d say its the single best tool we have to differentiate ourselves as the tide of slop rises - people.

There’s been much discussion in recent years that audiences have shifted somewhat in preference to not only value what they are being told, but who is telling them. There’s no doubt the dial of the “who” has shifted from general faith in news gathering organisations to that of individual journalists or content creators, whether they be part of an organisation or not. This is not something new really, and in the UK for example, you can argue it’s a return to our usual rancour and dispute after just a few decades of post-war consensus, hence the current struggles of the BBC.

This, I would argue, is personalisation in the shape of people choosing people to interpret the world for them. Publishers are full of people who can do just that.

---

Simplify audience management and first-party data, anywhere people and content meet.

Manage entitlements, craft engagement strategies, and model users and groups from one platform.

Glide Nexa goes beyond standard CDP functionality to help publishers foster community interaction, boost audience interaction, and build sustainable engagement strategies.

Request to see Glide Nexa in action.

---

Spain hands Meta a €479M media reality check
Stop me if you know this story: Meta is yet again in trouble for how it handles user data, and this time Spain is the one sending the bill. A court in Madrid has decided that Meta’s “creative” interpretation of GDPR, in order to get ahead in the ad race, resulted in nearly half a billion euros debt to local media outlets. The Spanish publishers claim that fiddling with the rules crushes their ad revenues, while Meta is shouting “nonsense” and gearing up for an appeal. It seems that collecting mountains of data under a legal excuse nobody bought was not the move.
Read

Reaching for the stars
Astrophysics analogies with user traffic data are fairly rare in publishing, but Piers North, CEO of one of the UK’s largest publishers, Reach, makes a good fist of it here: “There is no question a Big Bang has happened. And what you’ve seen is the universe has kind of massively expanded, but the number of stars has remained the same, and what has filled it is like dark matter. It’s this space that we’re not quite sure what is there. Is the traffic going elsewhere?”
Read

Italy rewrites Google’s playbook
Italy took one look at Google’s “Do you want to share your data?” screens and deemed them a bit too nudge-nudge-wink-wink, forcing the company to redesign them to be less confusing for the users. And they’re not stopping there, they’re eyeing Meta and WhatsApp as well. Even though the EU’s big digital law (the DMA) was supposed to centralise all of this in Brussels, national regulators like Italy keep taking the wheel from time to time, which doesn’t make Brussels angry, on the contrary, they’re excited, cheering from the sidelines, and planning to roll these changes out across the whole EU. Meanwhile, across the Atlantic the US is giving EU the side-eye for constantly smacking US tech giants with rules and fines. Italy followed in Germany’s footprints, who have done something similar, which prompts the questions whether the supposedly unified EU rulebook is actually turning into a patchwork of different (national) interventions.
Read

No to churnalism, yes to the authentic
Publishers are doubling down on authenticity, proving once again that humans still matter in content. The playbook is to treat reporters like brand ambassadors, create distinctive editorial voices, and make digital experiences feel premium. That will keep the readers loyal and make them come back for more, proving that even though AI is all around us, humans still call the shots. INMA knows more.
Read

Prospect for the win
Prospect Magazine clearly came to play at the PPA Independent Publisher Awards, snatching Commercial Leader of the Year as well as Newsletter of the Year, while the rest of the team bagged shortlists for Innovation, Independent Publisher, and Changemaker of the Year. Big cheers from their proud CMS supplier - that would be us - to the team for showing that UK indie media is thriving!
Read

Aussies are leading, not lagging
Australia isn’t taking AI safety lightly. In 2026, the government is launching the Australian Institute for Safe and Innovative AI, a place where new algorithms will be monitored, tested, and policed. Besides giving Australia a spot in the global AI safety network, it will help regulators keep pace with evolving AI and offer guidance to both businesses and the public, making sure that future “AI did what?” moments will be avoided.
Read

The return of Bread & Butter
There are more signs that legacy brands aren’t just content to fade into the archives: Consumer Reports in the US is dusting off Bread & Butter, a WWII-era shopping newsletter and serving it to a new generation, via Substack instead of via snail mail. While the 1941 version talked about nylon shortages, it’s younger brother is discussing chaotic online deals. The revived version is aimed at readers who probably only know Consumer Reports because their parents talked about it, pitching a weekly no-nonsense and hype-free shopping advice coming straight from their 90-year-old trustworthy friend.
Read

Is the end to AI scraping nigh?
The Internet Engineering Task Force has started to put shape to standards which might force AI companies to actually respect what site owners want to do with their content. The IETF’s AI Preferences Working Group is pushing for real and enforceable rules in robots.txt and HTTP readers. The plan is to have a universal vocabulary for AI permissions, a standard way to attach them to content, and a conflict-resolving system which would enable creators to throw AI crawlers out the door when they want to. Nothing is standardised yet, but if these stick, then AI scraping might get an expiration date.
Read

Artists in control
The music industry is not waiting around for the AI chaos to settle, as Warner Music and Suno have struck a deal to turn AI-generated music into a licensed operation instead of a free-for-all buffet. Things are moving quickly, as it was only last year that Warner was suing Suno for copyright violations of “an almost unimaginable scale”. Artists will get a say on how their names, voices, and songs are used, trying to put a stop to the era where AI models scrape first and ask questions never. In 2026, Suno will roll out new models, sunsetting free users while paid users will get their monthly fill. They are calling it a “pro-artist AI”, giving the musicians a chance to decide where their voices will end up.
Read

Chatbots VS chefs
AI took over the internet, and now it’s aiming for your kitchen. AI Overviews are dropping traffic for recipe writers, up to 80 per cent, and conveniently forgetting to link the original source. If you want to avoid burning or expiring from botulism, maybe it’s time to dust off grandma’s old cookbook instead of turning to a chatbot.
Read