When AI chatbots answer news-related questions, they play favorites — and the winners are a small handful of dominant publishers, according to new research from the UK.
Key Takeaways
- UK research shows AI chatbots heavily favor a handful of major publishers.
- BBC and Guardian get a disproportionate share of all AI citations.
- Smaller outlets are essentially invisible to AI-driven news discovery.
A study by British thinktank IPPR found that on average, 34 percent of journalistic citations from major AI tools go to just one news outlet. The most-cited source on each platform appeared four times more often than the second-place publisher.
The research, titled “AI’s Got News For You,” analyzed more than 2,500 links from ChatGPT, Google Gemini, Google AI Overviews and Perplexity in response to 100 news queries run in the UK market. The findings reveal how AI is reshaping which publishers get visibility — and which get squeezed out.
Google’s AI Overviews cited the BBC in 41 percent of news responses. The Guardian led on ChatGPT (27 percent) and Gemini (37 percent). Perplexity also surfaced BBC content most frequently — despite the broadcaster blocking its crawlers and threatening legal action last year.
The pattern matters beyond the UK. While the specific outlets differ by market, the underlying dynamic is universal: AI tools are creating a new layer of visibility that favors publishers with licensing deals or algorithmic preference, potentially locking out smaller and local news providers.
“Questions need answering around how financial relationships between AI companies and news brands shape AI answers,” the report states. “If licensed publications appear more prominently in AI answers, there is a risk of locking out smaller and local news providers who are less likely to get AI deals.”
The Guardian has a compensation agreement with OpenAI for ChatGPT citations. The BBC does not — and actively blocks AI crawlers. Yet Perplexity continues to surface BBC content, which the company attributes to third-party data partnerships. Infrastructure provider Cloudflare has accused Perplexity of using “stealth, undeclared crawlers” to bypass publisher restrictions.
The report recommends “nutrition labels” for AI-generated answers disclosing how sources are selected, along with collective licensing frameworks to help smaller publishers negotiate.
The UK Competition and Markets Authority said this week it would give Google 12 months to let publishers opt out of AI Overviews before considering action on payment terms. News Media Association chief Owen Meredith called on regulators to move faster: “Fair payment from the market leader is critical to a functioning licensing market.”
For US publishers watching from across the Atlantic, the question is whether American AI citation patterns show similar concentration — and whether anyone is measuring it.
