Google’s move to embed artificial intelligence more deeply into its search experience is having a profound impact on the digital publishing industry, where news organizations are now facing a noticeable collapse in the referral traffic they once relied on to sustain both audience engagement and advertising revenue.
What has traditionally been a system built on users clicking through to source material is gradually shifting towards a model where answers are delivered directly by algorithms that summaries content without necessarily crediting or linking to its origin, a development that has left many publishers scrambling to adapt. The latest data, compiled by analytics firm SimilarWeb and cited in a Wall Street Journal investigation, confirms the severity of the disruption, with HuffPost having lost more than half its traffic over a three-year period and Business Insider reporting a similar 55 percent decline that coincided with the recent decision to cut a fifth of its workforce.
The New York Times, while maintaining its position as a leading digital news provider, saw the proportion of its search-driven traffic drop to 36.5 percent in April 2025 from nearly 44 percent three years earlier. The Washington Post followed a similar trajectory, and although The Wall Street Journal recorded an increase in the raw number of search visits, its reliance on that stream weakened, with its share of total traffic falling to 24 percent from a previous 29 percent.
The changes began taking root with the introduction of Google’s AI Overviews, a tool that began providing brief responses to user queries by drawing on content across the web and presenting it directly within search results, reducing the need to click through. With the arrival of Google’s AI Mode, a chatbot-like experience similar to OpenAI’s ChatGPT, the effect is expected to become even more pronounced since fewer links are included in its conversational outputs and users are encouraged to stay within Google’s environment.
While the consequences are most visible across newsrooms, the ripple effects are spreading well beyond journalism, with sectors like health information, product reviews and travel guidance all reporting sharp declines in traffic as AI tools present information that once drew readers to specialized sites. At the same time, concerns are mounting over how much of the content used to generate these AI responses originates from publishers without any formal arrangement, prompting some media companies to seek licensing deals as a defensive measure.
The New York Times recently entered into an agreement with Amazon to supply editorial material for artificial intelligence development, a move that follows similar deals involving OpenAI and publications such as The Atlantic. Meanwhile, AI start-up Perplexity has introduced a proposal to share advertising revenue with news providers when their work is surfaced in automated responses, though how widely such models will be adopted remains uncertain.
Inside publishing circles, there is growing acknowledgement that the familiar pathways of online visibility are changing quickly and that dependency on traditional search is no longer a reliable strategy. With algorithmic systems increasingly determining what gets seen and what does not, editorial leaders are confronting a moment that demands not just short-term adjustments but deeper shifts in how journalism connects with audiences in an AI-dominated landscape.
Image: DIW-Aigen
Read next:
• YouTube’s Expanding Creator Network Fuels a Booming Digital Economy in the U.S.
• How Many People Visit a Website? These 6 Free Tools (With Paid Features) Can Help You Analyze That
What has traditionally been a system built on users clicking through to source material is gradually shifting towards a model where answers are delivered directly by algorithms that summaries content without necessarily crediting or linking to its origin, a development that has left many publishers scrambling to adapt. The latest data, compiled by analytics firm SimilarWeb and cited in a Wall Street Journal investigation, confirms the severity of the disruption, with HuffPost having lost more than half its traffic over a three-year period and Business Insider reporting a similar 55 percent decline that coincided with the recent decision to cut a fifth of its workforce.
The New York Times, while maintaining its position as a leading digital news provider, saw the proportion of its search-driven traffic drop to 36.5 percent in April 2025 from nearly 44 percent three years earlier. The Washington Post followed a similar trajectory, and although The Wall Street Journal recorded an increase in the raw number of search visits, its reliance on that stream weakened, with its share of total traffic falling to 24 percent from a previous 29 percent.
The changes began taking root with the introduction of Google’s AI Overviews, a tool that began providing brief responses to user queries by drawing on content across the web and presenting it directly within search results, reducing the need to click through. With the arrival of Google’s AI Mode, a chatbot-like experience similar to OpenAI’s ChatGPT, the effect is expected to become even more pronounced since fewer links are included in its conversational outputs and users are encouraged to stay within Google’s environment.
While the consequences are most visible across newsrooms, the ripple effects are spreading well beyond journalism, with sectors like health information, product reviews and travel guidance all reporting sharp declines in traffic as AI tools present information that once drew readers to specialized sites. At the same time, concerns are mounting over how much of the content used to generate these AI responses originates from publishers without any formal arrangement, prompting some media companies to seek licensing deals as a defensive measure.
The New York Times recently entered into an agreement with Amazon to supply editorial material for artificial intelligence development, a move that follows similar deals involving OpenAI and publications such as The Atlantic. Meanwhile, AI start-up Perplexity has introduced a proposal to share advertising revenue with news providers when their work is surfaced in automated responses, though how widely such models will be adopted remains uncertain.
Inside publishing circles, there is growing acknowledgement that the familiar pathways of online visibility are changing quickly and that dependency on traditional search is no longer a reliable strategy. With algorithmic systems increasingly determining what gets seen and what does not, editorial leaders are confronting a moment that demands not just short-term adjustments but deeper shifts in how journalism connects with audiences in an AI-dominated landscape.
Read next:
• YouTube’s Expanding Creator Network Fuels a Booming Digital Economy in the U.S.
• How Many People Visit a Website? These 6 Free Tools (With Paid Features) Can Help You Analyze That
