Is Personalised News a Brave New Future – or Just Another Filter Bubble?

The BBC is turning to AI to personalise news. The Independent is using it to generate custom summaries.

On the face of it, this feels like a logical next step in how people consume information—tailored to their interests, in digestible formats, delivered at speed. AI is a natural tool to support this. 

And there are clearly upsides. For one, it could help counter the growing trend of news avoidance. People are tuning out of traditional news because they feel overwhelmed, disconnected—or, quite simply, exhausted by the never-ending cycle of depressing headlines. Personalisation promises to make news feel more relevant, more manageable, and more likely to connect. 

But it also raises bigger questions. The strength of traditional journalism—at least in principle—was that it presented a broader view of the world. A newspaper, a TV bulletin, a radio report: they all offered stories you might not seek out yourself, but which were nonetheless important. When the news becomes hyper-personalised, that serendipity is lost. We risk ending up in the same place we’ve seen social media take us—fragmented echo chambers where people are only exposed to what they already know or already care about. In that scenario, balanced perspectives and public debate suffer. 

Let’s also not pretend traditional media was ever entirely neutral. Longstanding allegiances between publishers and political parties have always shaped coverage. But at least the editorial process was visible—and shared. When AI becomes the gatekeeper, there’s less transparency, more opacity, and new risks. Apple recently paused the use of AI in its news app after it published error-ridden summaries—proof that this technology still isn’t ready for editorial prime time. 

That said, it’s possible to imagine more optimistic outcomes. If you’ve shown interest in humanitarian crises like Gaza, for instance, the tech could serve you stories about similar—but less widely reported—conflicts, such as Sudan. In theory, it could challenge the biases of traditional editorial decisions, not just reinforce them. 

But this feels more like an exception than a rule. If our experience with social media is anything to go by, AI-led personalisation will likely reinforce preferences, not stretch them. And that risks making our information landscape narrower, not richer. 

So, what does this mean for PR? 

On one level, it plays to our strengths. We’ve always built campaigns around clearly defined audiences. But in a world where two people searching the same thing may get completely different results, channel mastery becomes critical. We’ll need a sharper understanding of how and where earned media is being surfaced—and to whom. 

The opportunity is real. But the risks are growing. PR needs to be ready for both. 

By Ricky Ambury, Senior Director at Citypress