Ask an AI chatbot what’s happening in the world, and chances are it’s already decided which voices you’ll hear from.
New research suggests that decision is far from neutral.
According to a fresh analysis, popular AI tools increasingly rely on a limited set of news outlets — with a noticeable tilt toward left-leaning sources.
Think-Tank Warns of a Narrow Information Diet
The findings come from the Institute for Public Policy Research, which examined how major AI platforms answer questions about current events.
Their conclusion was blunt: the pool of news sources being used is both narrow and inconsistent, raising concerns about how balanced — or unbalanced — the information landscape is becoming.
BBC and Guardian Dominate AI Answers
When researchers dug into the data, clear patterns emerged.
Google’s AI Overview leaned heavily on the BBC, citing it in just over half of its answers.
Meanwhile, ChatGPT and Google Gemini most frequently turned to the Guardian, which appeared in more than half of their responses.
Right-Leaning Outlets Barely Make the Cut
In contrast, publications such as the Daily Mail and other right-leaning outlets showed up far less often.
That’s despite the fact that these organisations collectively reach a significant share of the UK news audience.
The imbalance has sparked fears that AI tools are quietly reshaping which publishers win attention — and which ones lose out.
A New Set of Winners and Losers Emerges
IPPR researchers say this trend risks creating a media ecosystem where some outlets are amplified by default, while others are effectively sidelined.
Over time, that could limit the range of viewpoints users encounter and subtly push particular narratives to the forefront.
Who Decides What AI Can Read?
Behind the scenes, AI companies make complex choices about which publishers to pay, prioritise, scrape for free, or exclude entirely.
Some outlets, including the Guardian, have licensing agreements with AI firms.
Others, like the BBC, have attempted to block AI access altogether.
Yet users are rarely told how or why these sourcing decisions are made.
The Transparency Problem Users Never See
One of the biggest red flags raised by researchers is the lack of visibility.
When an AI chatbot delivers an answer, there’s no clear explanation of what was left out — or why.
That hidden editorial process means readers often assume neutrality where none may exist.
The Financial Squeeze on Journalism
Beyond questions of balance, there’s a money issue looming large.
Studies show that when Google’s AI Overview appears in search results, users are nearly half as likely to click through to news websites.
Fewer clicks mean shrinking ad revenue and fewer subscriptions — a dangerous mix for already strained newsrooms.
Publishers Push Back Against Big Tech
Industry leaders are now urging regulators to step in.
News Media Association chief executive Owen Meredith warned that weakening copyright protections would strip publishers of fair compensation for the journalism AI systems rely on.
He argued that regulators must stop tech giants from using their market power to extract content without payment.
Calls Grow for Government to Step In
IPPR experts say the moment has arrived for policymakers to take a firmer grip.
While early AI policy focused on speeding up innovation, they argue the priority now should be shaping technology to serve the public good.
In the news world, that means protecting diversity, ensuring fair payment, and preventing AI from quietly hollowing out the information ecosystem.
A Crossroads for AI and the Public Sphere
The message from researchers is clear: the tools already exist to make AI a force for better journalism rather than weaker journalism.
But without government action and stronger oversight, they warn the current trajectory could narrow public debate — and lock in long-term damage before anyone fully realises what’s been lost.
Share on Facebook «||» Share on Twitter «||» Share on Reddit «||» Share on LinkedIn