How AI Search Tools Are Worrying Online Publishers

The relationship between search engines and news publishers has always been uneasy, but it worked. It was for two decades that sites like Google acted as the gateway, sending millions of readers toward articles, features, opinion pieces, and reviews. Publishers built whole strategies around headline optimisation and keywords, confident that ranking well meant reliable traffic and revenue. That balance, however, is now changing.

In as little as the past year, Google has rolled out AI Overviews and an AI Mode tab. Instead of a traditional list of links, users are met with chatbot-style summaries that often answer the question without requiring a click-through. For news organisations already under pressure from falling advertising income, rising costs, and declining print circulation, this feels like a new kind of threat. Some publishers have reported steep drops in referral traffic, raising the fear of a “zero-click” future. 

Industry groups in the UK, including the owners of leading newspapers and magazines, have already called on regulators to demand more transparency. Their argument is straightforward: if Google’s AI is reproducing their work in summary form, they should know how much traffic is being lost and what agreements can be reached to avoid being sidelined. Without that clarity, smaller outlets in particular risk becoming invisible.

AI as a Tool

Much of the conversation so far has been framed around what publishers stand to lose. But there is another side to the story. AI is already extraordinary valuable in many creative and entertainment sectors, and it often complements human work rather than replaces it in any form.

One good example comes from online casinos, where AI has been used to enhance the likes of slot games. Escapist Magazine analysis suggests that iGamers are particularly drawn to slots because of their high RTP, ease of play and the vast number of slot games available. By analysing patterns of play, AI helps developers design more engaging features, tailor difficulty, and create personalised recommendations for players. Human attempts at this would cost time and money, two things which any self-respecting casino wants to save on at all costs. 

Publishing and media can look to other industries for lessons, too. The giant streaming platforms rely heavily on AI recommendation systems, and the results are clear: Netflix has said that over 80% of the shows watched on its service are discovered through its AI-driven suggestions, rather than manual browsing. This means audiences are exposed to new films and series they’d maybe never have searched for, so they stay engaged longer, and smaller productions get the advertising they have always dreamed of. 

In the gaming industry, developers are using AI to create adaptive non-player characters (NPCs) that behave differently depending on how players interact with them. A 2024 survey by the Game Developers Conference found that 31% of studios are already experimenting with AI for character design and world-building. This gives players less predictable experiences and adds replay value; something traditional programming would have really struggled to achieve at scale.

Even in music, AI is making its mark in the production process. Tools for automated mastering, such as LANDR, are used by over 5 million artists worldwide, according to company figures. These systems work by handling repetitive technical adjustments quickly, so musicians and artists are free focus on their writing and performance. It’s not about replacing creativity because they provide a cost-effective way for independent artists to get professional-quality sound without expensive studio time that would previously have been required.

AI doesn’t have to strip away any value here, because when its used well and responsibly, it can expand the reach of creative work and increase audience engagement, the latter of which is, bluntly speaking, the ultimate goal.

The Pressure on News Outlets

Still, it would be naïve of us to ignore the risks for the publishing industry. Some news outlets have reported referral drops of more than 70% since AI summaries began appearing more prominently. If a machine-generated overview provides the basic facts, readers may never click through to the original article. The result? Fewer impressions for advertisers and fewer opportunities to convert casual visitors into subscribers.

The issue isn’t just financial. AI systems are known to produce inaccuracies, called hallucinations. Early versions even suggested dangerous or absurd advice, though most of those errors have since been corrected. Even so, the possibility of misinformation remains, particularly if summaries condense complex investigations into a handful of sentences without context. For journalism, which relies on carefully constructed nuance, this is a real problem.

There’s also, understandably, concern about fairness. When Google or Apple present AI-generated snippets, they are drawing from published material that has taken time, money and genuine effort to produce. Publishers argue that this is effectively a free extraction of value. Unless rules are put in place to require permission or compensation, many fear that the intellectual effort of the media industry will be absorbed into vast training datasets with little return.

Moving Patterns

Algorithm-driven feeds are complicating patterns. Google Discover, which pushes articles to users based on their browsing habits, is becoming a bigger and bigger source of traffic than the traditional search. But this traffic is unpredictable. It often favours punchy or sensational headlines rather than long-form investigative pieces. For those publishers that are trying to maintain high editorial standards, building a strategy around Discover can feel like chasing a moving target.

Analysts point out that these feeds are not central to Google’s core business. If they are used primarily to offset criticism about declining search referrals, publishers risk becoming dependent on a channel that may not have long-term stability. And that instability makes it difficult for outlets to plan investments in reporting teams, data projects, or new sections.

The Fight Over Content Rights

Behind the traffic concerns, there’s a broader question of ownership. AI companies need vast amounts of data to train their systems, and much of that data comes from published writing. Creative industries, including news media, are lobbying governments to ensure that copyright is respected. Their fear is that without legal safeguards, value will be siphoned away from a multi-billion pound industry without fair compensation.

This tension is not going away soon. AI has shown that it can be useful in entertainment and creative spaces when implemented responsibly, from enhancing games and music to making media discovery easier. But when the same technology undermines the very structure that funds professional reporting, the result is a clash of priorities. Publishers know that innovation cannot be stopped, but they also know that journalism cannot survive if its economic base is eroded.

A Turning Point?

Make no doubt about it: the next few years will be decisive. Regulators are weighing how to balance competition, transparency, and innovation. Publishers are testing new ways of working with AI, from licensing deals to building their own tools. And audiences are adjusting to a world where answers appear instantly, without needing to browse through multiple sites. The challenge is to ensure that quality reporting continues to be valued. AI can play a role in supporting creativity, but only if the balance between technology companies and content producers is handled fairly. 

Otherwise, the very institutions that are supposed to provide us with trusted information risk being weakened at a time when, arguably, we need them most.