Google once promised to organize the world’s information. Now its AI tools risk burying it.
The New York Times reported this week that the company’s aggressive push into generative search features has accelerated a years-long decline in traffic to independent websites. Publishers watch visitors evaporate. Writers question their futures. And users receive neatly packaged answers without ever leaving Google’s domain.
But the story runs deeper than vanishing clicks. It reveals a fundamental clash between two visions of the internet. One values discovery through links and exploration. The other prioritizes speed and synthesized responses. Google insists both can coexist. Evidence suggests otherwise.
At the heart of the conflict sits AI Mode, the conversational interface Google rolled out globally after its I/O event in May. Powered by the new Gemini 3.5 Flash model, it transforms the familiar search box into something closer to a personal assistant. Type a query. Receive a custom summary. Follow up in chat. Rarely click through.
Elizabeth Reid, Google’s vice president of search, called the changes “a new era for AI Search” in the company’s official blog post. The upgrade marks the biggest overhaul to the search box in more than 25 years. Users can now upload images, videos or even Chrome tabs as inputs. Agents handle complex tasks across the web.
Yet that convenience comes at a cost. Zero-click searches have climbed past 60 percent, according to analyses cited by The Next Web. News sites report double-digit traffic drops. Some see worse.
Data from Similarweb, referenced in NPR’s coverage last year, showed CNN traffic down 30 percent, Business Insider and HuffPost around 40 percent. Those trends have only sharpened. A large lifestyle publisher modeled scenarios for Digiday that forecast Google referrals shrinking to single digits within years if AI Mode expands unchecked.
Chartbeat data shared with Press Gazette documented a 33 percent fall in search engine referrals across more than 2,500 news sites globally. Ahrefs measured a 58 percent drop in click-through rates for top-ranking pages when AI Overviews appear. The numbers paint a stark picture. Traffic that once flowed outward now pools inside Google.
Publishers aren’t passive. Many have adapted their content strategies. Some optimize for AI summaries by writing clearer, more structured material. Others focus on brand loyalty and direct visits. A few explore subscriptions or alternative platforms.
Still, frustration mounts. Nick Kramer, a digital strategist quoted in recent discussions, captured the mood. “The open web is on its way out. With A.I., Google is reducing everyone to raw data providers.” His words echoed across LinkedIn and industry forums this spring.
Publishers push back as Google doubles down
Google counters that its features send billions of clicks to the web each week. A spokesman told the New York Times that AI tools highlight links and meet evolving user preferences. The company introduced new Search Console reports showing impressions in AI Overviews. It expanded Preferred Sources signals to reward trusted publishers.
But those measures feel incremental. Core updates in May 2026 aimed to surface higher-quality content. Traffic reports remain opaque on actual clicks from AI interfaces. Publishers want transparency. They want data on which summaries drive visits and which replace them.
Glenn Gabe, an SEO consultant, highlighted the gap on X today. He noted Google needs to add click metrics to AI performance reporting. Without that visibility, publishers operate in the dark.
And the darkness spreads. Small businesses, recipe sites, how-to guides—all once thrived on Google’s referrals. Now AI condenses their expertise into paragraphs that satisfy curiosity without credit or compensation. Some creators report 70 to 90 percent drops in organic search visits.
The New York Times article details how this shift echoes earlier warnings. Back in 2023, when Google first tested Search Generative Experience, experts predicted exactly this outcome. Years later, the predictions hold. Only the scale has grown.
Regulators take notice. The European Union ordered Google to open Android features to rival AI assistants and share anonymized search data starting January 2027. Posts on X this morning buzzed with the news. One analyst called it regulators doing “what Google wouldn’t do on its own.” The move aims to loosen the company’s grip on discovery.
Yet Europe’s action addresses only part of the problem. The deeper issue lies in incentives. Google’s business model rewards engagement inside its properties. AI summaries keep users longer. They reduce the need to visit third-party sites. Revenue from ads on those summaries stays with Google.
Some voices argue this evolution was inevitable. Search has always mediated the web. As capabilities advanced from keywords to natural language to full conversation, the intermediary grew more powerful. But power without accountability breeds distortion.
Consider the information quality question. AI models trained on web data can propagate errors or flatten nuance. Publishers who invest in original reporting see their work repurposed without context. Readers receive answers stripped of bylines, sources or competing viewpoints. Trust erodes quietly.
Google says it cites sources. Links appear. But studies show click rates plummet when summaries precede results. Users stop at the answer. The link becomes decoration.
So what happens next? Some publishers bet on email lists, communities and products. Others court social platforms or emerging AI search tools like Perplexity. A few experiment with their own agents that pull from diverse sources.
Google itself faces pressure to evolve. Its May 2026 core update and new reporting tools signal awareness. Executives talk about partnership. Yet the product roadmap points toward more AI, more summarization, more containment.
The open web’s defenders point to history. The internet flourished because information flowed freely. Barriers to entry stayed low. Discovery rewarded quality and originality. If that loop breaks—if creation yields diminishing returns—then quality itself suffers.
Industry watchers on X today debated the timeline. Some predict a 30 to 50 percent further drop in referral traffic by 2028. Others see a floor where curiosity or complex needs still drive clicks. No consensus exists. Data remains fragmented.
One thing feels clear. The companies that adapt fastest will shape what comes after. Those who treat the web as raw material for AI training may find fewer humans willing to produce it. Those who build bridges back to original sources may earn loyalty.
Google holds the dominant position. Its decisions carry outsized weight. The question isn’t whether AI belongs in search. It does. The question is whether search can preserve the web that feeds it.
So far, the balance tilts one way. Publishers feel the strain. Readers enjoy convenience. And the open web, that messy, vital network of independent voices, shrinks a little more each quarter.
Change arrives faster than adaptation. Google’s latest moves only accelerate what began years ago. The real test lies ahead. Can the web sustain itself when its biggest gateway no longer needs to send traffic through the gate?
WebProNews is an iEntry Publication