The scale of generative artificial intelligence's penetration into the web environment turns out to be far more serious than commonly assumed. My analysis of recent data shows that about 10% of all existing web pages bear clear traces of synthetic authorship. However, this is merely the tip of the iceberg—among content published after ChatGPT's debut, the share of AI-generated texts exceeds one-third.
The New Reality of the Digital Landscape
As part of a large-scale study, a random sample of 10,000 English-language pages collected in July 2026 via the Common Crawl archive was analyzed. For identification, a machine learning model was used, tuned to detect linguistic patterns characteristic of modern neural networks.
The figure of 10% may seem moderate, but it is deceptive. The sample included a vast layer of legacy content created long before the era of generative models. If we filter out the historical layer and focus exclusively on materials published after November 2022, the picture changes radically: more than 33% of new pages show signs of AI generation.
The Commercial Sector—the Main Area of Impact
The distribution of synthetic content across domain zones is extremely uneven and highly telling. The commercial .com segment leads, where every tenth page contains AI traces—twice the rate of .org (4.6%). Educational (.edu) and governmental (.gov) resources turned out to be ten times cleaner—around 1%.
This disparity is explainable: business websites actively automate the production of product descriptions, SEO articles, and news briefs. These are voluminous, template-based formats that are ideally suited to algorithmization. Notably, at the dawn of ChatGPT's spread, the gap between zones was minimal—now we are witnessing rapid polarization.
Methodological Nuances and Conclusions
It is important to emphasize: the study is not an exact count, but rather a probabilistic estimate. The authors did not check edit histories or survey webmasters. The analysis relied on statistical linguistic markers, so some pages may have been written by humans but stylistically imitated AI, or conversely—undergone minimal editing without falling into the category of "substantially modified."
Nevertheless, the trend is obvious: we are witnessing a tectonic shift in the production of internet content. Notably, against this backdrop, Anthropic has already introduced mandatory labeling for its Claude models—a step that will likely become an industry standard.
My comment: The market is on the verge of an "inflationary collapse" of content. When a third of new pages are generated by machines, the value of textual information as such drops sharply, and trust and source verification come to the forefront. This is a fundamental challenge for the SEO industry and media, requiring a rethink of the very principles of ranking and copyright.