From Saved Recipes to Sinister Tech
Picture this: you’re enjoying a lovely morning scroll, swiping through your Insta explore page, when you spot a recipe you love (instant save for later), hear a song you were just raving about to a friend, and watch a workout reel that was exactly the motivation you needed. Maybe you were even debating deleting Instagram, but now you’re not so sure. After all, it knows you too well…
Most of the time we don’t question it. But eventually you think: I wonder how Instagram knows that I joined the gym last month, or that one of my best friends just got pregnant?
Hyper-personalisation seems harmless (albeit a bit spooky), but a recent BBC article pointed out a sinister side to social media algorithms. Women whose algorithms picked up that they were pregnant saw their feeds explode with wholesome baby content and baby-related ads, but when they lost their babies, the adverts didn’t stop (BBC).
💬 One of the women told the BBC: “Technology doesn’t understand loss and in moments when we least expect it, it reminds us with devastating precision of what we no longer have.”
After Tanya O’Carroll took her case to Meta alongside a legal team, Meta agreed to stop direct marketing (aka invasive ads), and Tanya is now thought to be the only UK Facebook user who is not targeted with personalised content (BBC). More than 10,000 people have since raised an objection to Meta to stop using their data for direct marketing, meaning Zuckerberg may be in for more lawsuits.
“Consent or Pay”
In September last year, Meta said they’ll introduce a subscription service for users who don’t want to see targeted ads on Facebook and Instagram, starting from £2.99 per month for UK users (BBC). In the past week that change has officially come into effect.
Some websites have adopted a similar model for users who don’t want cookies, with “Reject and Pay” buttons appearing under “Accept and Continue” (Data Protection Network).
Question is, should we really have to pay to stop targeted ads, or to reject cookies? If you’re like me then you can *usually* be bothered to follow the “Do Not Consent” button, but paying to reject cookies? Forget it.
Hyper-Personalisation in AI
Beyond social media, we should expect to see hyper-personalisation everywhere.
A few weeks ago I joined a webinar where AI experts were asked: “What are you most excited about for the future of AI?” The answer was personalisation. They predicted chatbots becoming more and more personalised to you, the questions you ask, and the way you like to chat.
OpenAI took a step (albeit a very small one) towards this with the launch of their GPT-5.1 (OpenAI), which allows users to choose tone of voice.
But what would it look like for ChatGPT to know you as well as your For You Page does? Imagine a virtual assistant that remembers your preferences, your goals, your tone, your bad habits. Like algorithms, it’s exciting and a little eerie. There’ll be benefits we haven’t thought of yet, but – learning from the story above – some consequences too.
How Important is AI Advertising to Marketers?
OpenAI have confirmed that in 2026 they’ll be integrating ads directly into GPT chats, beginning with non-paying users (NP Digital). This means that in 10 or so years, AI advertising could be a big part of paid media strategy. AI visibility already changes by the day, but with paid ads involved, things could become even more volatile. So where should marketers put their money?
On the one hand, marketers who want to stay ahead need to optimise content for AI visibility, and plan for future ad integration (HubSpot). On the other hand, there’s a growing feeling that we might be getting ahead of ourselves.
During an Ahrefs podcast, SEO legend Patrick Stox revealed that, in reality, Large Language Models (LLMs) account for less than 1% of online traffic (Ahrefs). A study of 1.5 billion searches found that Google accounts for 34.5% of total search share, followed by YouTube (24%), TikTok (16.7%), and Instagram (20.9%).
Following Google then, Youtube is actually the second largest search engine. So before you fall into the trap of spending all your time optimising for generative search, trust that you know your audience and take a closer look at where they are. That may be ChatGPT, but it’s more likely Google, Youtube, TikTok, or Instagram. So don’t let your social media strategy suffer.
If you’re looking for a social media agency who just gets it, head to Our Own Brand (OOB). We’re independent, plugged in, and known for making noise. Trusted by global brands. Built for now. Obsessed with what’s next.
Drop us a message! hello@ourownbrand.co
Conclusion
Algorithms don’t just guess what we want – they know. From personalised ads to AI companions that mirror our moods and habits, hyper-personalisation is becoming the norm. But as we’ve seen, this precision comes with a price: privacy, emotional sensitivity, and the burden of consent shifting to the user.
Whether it’s a grieving mother bombarded with baby ads or a chatbot that knows you a little too well, the question isn’t just about what algorithms can do, it’s about what they should do.
As marketers, technologists, and users, we need to decide where the line is. AI advertising may be the next frontier, but it must be built on trust, transparency, and user control, not just clever targeting.
The algorithms are getting smarter. It’s up to us to make them more humane.
This one's on us.
OOB is a social media agency. If this sparked something, let's talk.