Google is promoting the rapid adoption of its artificial intelligence features, while new research and reports from publishers raise concerns about the technology’s impact on web traffic and the quality of information. The conflicting narratives highlight a central tension in the current evolution of web search.
On a Q2 2025 investor call, Alphabet and Google CEO Sundar Pichai announced significant growth for the company’s consumer-facing AI tools. According to Pichai, AI Overviews—the feature that provides AI-generated summaries at the top of search results—now serves 2 billion monthly users across 200 countries, an increase from 1.5 billion in May 2025. Other AI products also show strong growth, with the Gemini app reaching 450 million monthly active users and the AI Mode in Search acquiring 100 million monthly users in the U.S. and India. Google states that AI Overviews are driving over 10% more search queries for the types of searches that display them, suggesting increased user engagement.
Study shows AI Overviews reduce web traffic
In contrast to Google’s positive framing, a study from the Pew Research Center presents a different picture. The analysis, based on data from 900 users in March 2025, found that the presence of an AI Overview significantly reduces the likelihood that a user will click on a traditional web link.
- Searches without an AI Overview had a click-through rate of 15%.
- Searches with an AI Overview saw this rate drop to 8%.
- Critically, only 1% of searches resulted in a user clicking on one of the source links cited within the AI Overview itself.
The study also noted that AI Overviews appear in about one out of every five searches, with a much higher frequency for searches phrased as questions (60%). This suggests that for many users, the search journey ends with the AI-generated summary, cutting off traffic to the websites that supply the underlying information.
The effect on publishers and information quality
Web publishers argue this trend is creating a “traffic apocalypse.” The tech publication 404 Media provided a concrete example: after publishing an original story on AI-generated music, it found that Google’s AI Overview summarized its reporting but linked to aggregator sites that had copied the story, not to the original source. This practice deprives original content creators of traffic, which is essential for their business models, whether ad-based or subscription-driven.
Beyond the economic impact, critics point to significant issues with the reliability of AI-generated answers. Generative AI is prone to “hallucinations,” or presenting false information as fact. Well-known examples include AI Overviews advising users to add glue to pizza, based on a joke from a Reddit forum, and incorrectly stating that the living author Dave Barry was deceased.
The system has also proven vulnerable to deliberate manipulation. Artist Eduardo Valdés-Hevia reported that he was able to trick AI Overviews into presenting his fictional horror concepts, such as the “Parasitic Encephalization Theory,” as real scientific theories. He and others were able to get Google’s AI to report on a fake phenomenon called “AI Engorgement” within hours of posting about it online, demonstrating how easily coordinated disinformation campaigns could exploit the feature.
In response to the Pew study, a Google spokesperson stated that the research “uses a flawed methodology and skewed queryset that is not representative of Search traffic.” The company maintains that it has “not observed significant drops in aggregate web traffic” and that its AI features create “new opportunities for people to connect with websites.”
Sources: TechCrunch, Ars Technica, 404 Media