Will personalised news narrow our minds not broaden them?

If a publication is telling you one thing and your neighbour something else, is it still in the business of news?

We've come to accept customised search results, targeted online adverts and filtered Facebook feeds, but now media companies are getting in on the act.

For some time now The Mirror website has been serving up stories based on what they call "activity data" and "social data insights from Facebook”. What that means is that the selection of stories you see is determined in part by what you've clicked on before and what Facebook has deduced about you and your interests.

Subscribe to The Week

Escape your echo chamber. Get the facts behind the news, plus analysis from multiple perspectives.

SUBSCRIBE & SAVE

Sign up for The Week's Free Newsletters

From our morning news briefing to a weekly Good News Newsletter, get the best of The Week delivered directly to your inbox.

From our morning news briefing to a weekly Good News Newsletter, get the best of The Week delivered directly to your inbox.

Sign up

The attraction for the publisher is obvious: if I'm presented with stories tailored to my interests, I'm more likely to read them, more likely to share them with like-minded friends and more likely to return.

That translates into increased revenue, especially with advertisers willing to pay a premium to reach certain niche audiences. A reader interested in house prices and mortgage rates is highly prized by banks – and therefore also to those selling advertising to banks.

For the reader, too, the advantages are apparent: less time wading through stories of little interest and more time to enjoy the others.

As the technology develops, personalisation will become more sophisticated. News apps that can process the GPS data in your phone could prioritise news in a highly personal way: regular rail commuters, for example, might see more stories about plans for high-speed rail. Frequent visitors to Heathrow might get features on airport expansion instead, as well as more adverts from airlines.

The disadvantages, however, could be even more profound.

The examples above, though untroubling in themselves, hint at the way in which personalised news could work against social cohesion. News outlets would no longer address a single audience, but a shifting mass of interest groups whose loyalty would depend on a regular stream of stories that reflected and confirmed their own world view.

Pitting rail enthusiasts against plane spotters may not undermine society, but what if publishers discovered that they could win over readers by telling different stories depending on the wealth of the reader, or the religion or race?

Already, it could be argued, there is a great deal of self-sorting: open the Daily Mail and The Guardian and you will find the same story told in starkly different terms.

But both of those publications are open for all to see. Stories can be challenged, overturned and corrected. Individual readers – those who care – can guard against bias by looking at more than one source.

If, say, I'm suspicious of the way The Guardian is reporting a story, I can click across to The Telegraph, The Times and the BBC. And if I read the Daily Mail, that is a choice I have made, not one that has been made for me.

The invisible hand of the algorithm would be more insidious – and much harder to correct. If The Guardian thinks I want to read about only the benefits of High Speed 2, then The Times, looking at the same data, would probably think the same.

Instead of choosing between other people's biases, I would have my own reflected back to me – and in such a way that they seemed universal, even if they were unique to me.

The consequences could be far-reaching. In Rotherham, for example, the widespread abuse of children by men of Pakistani descent was exposed only because The Times worked for years to bring it to the attention of those in the centre-ground of politics, many of whom were resistant to the idea that the problem was real.

If those stories had been served up only to those who were predisposed to believe them – those who will believe any negative story about immigrants or Muslims or anyone else who is not like them – and not to those who were sceptical, the campaign would have been fatally undermined.

Or consider this summer's war in Gaza. If a news algorithm had detected that a reader's sympathies lay with one side or the other and then pandered to them, however subtly, a bad situation would be made yet worse.

The internet once promised to break down barriers by letting us see the world from any perspective we chose, but many of us have chosen to stick with just the one. Soon those choices will be reinforced by news delivery systems in subtle ways that we may neither understand nor perhaps even notice.

Arthur Miller said a good newspaper was a nation talking to itself. The danger of personalised news is that the national conversation splinters into a clamour of individual voices, muttering to themselves while no one else is listening.