Pew Finds AI Text Surged on Web After ChatGPT Launch
A Pew Research Center analysis reveals that over a third of web pages published since ChatGPT's launch show signs of AI generation, marking a massive shift in how online content is created.

The Pew Research Center analyzed nearly half a million English-language web pages from the Common Crawl web archive to track the rise of machine-generated content. Using the Open Pangram AI detection tool, researchers discovered that more than a third of pages published after the late 2022 launch of OpenAI's ChatGPT show signs of AI authorship. In a broader sample from July 2026, approximately 10 percent of all analyzed pages displayed clear signs of machine writing, illustrating a steady upward trajectory.
The distribution of this automated text is highly uneven across the web. Commercial ".com" domains are roughly ten times more likely to feature AI-written text than educational or government sites. Specifically, about 10 percent of ".com" pages showed signs of AI authorship, compared to 4.6 percent of ".org" domains and just 1 percent for both ".edu" and ".gov" sites. Stylistically, the study noted a massive spike in AI-favored vocabulary and punctuation. Since 2023, the use of em dashes has doubled, Oxford comma usage has risen by 63 percent, and words like "delve," "interplay," "testament," and "tapestry" have more than doubled in frequency. Additionally, "it's not just X, it's Y" phrasing nearly tripled on the web and quadrupled in corporate PR documents.
These findings align with an April 2026 study by Imperial College London, the Internet Archive, and Stanford University, which also found that roughly 35 percent of newly published websites were at least partially AI-generated, exhibiting a 33 percent higher semantic similarity and a more positive tone. However, Pew acknowledged that tools like Open Pangram have significant limitations. They struggle to differentiate between fully automated text and human writing that has merely been polished with AI assistance, and they frequently misfire.
For digital marketers, SEO specialists, and content creators, this saturation means that generic, AI-style writing is rapidly losing its distinctiveness. As the web becomes crowded with repetitive vocabulary and predictable syntax, practitioners must focus on developing highly unique human perspectives or risk their content blending into the background. Furthermore, the stigma surrounding AI usage means creators must navigate a polarizing landscape where over-reliance on automated drafts can damage brand credibility.
This is our own summary of reporting by The Decoder



