Newsletter Subscribe
Enter your email address below and subscribe to our newsletter
[forminator_form id="25163"]
All voices matter
Pew study: A third of post-ChatGPT webpages show AI authorship
- Pew Research Center found that 35% of webpages published since ChatGPT's November 2022 launch show significant signs of AI authorship, based on analysis of Common Crawl archives.
pewresearch - AI-typical linguistic markers — including em dashes, Oxford commas, and words like "delve" and "tapestry" — have roughly doubled across the web since early 2023, according to the study.
pewresearch - The report follows Cloudflare's finding that bot traffic now exceeds human traffic online, underscoring a broader shift in how web content is produced and consumed.
techcrunch
Over one-third of English-language webpages published since ChatGPT's November 2022 release show signs of AI authorship, according to a study released Thursday by Pew Research Center.pewresearch
The finding offers one of the clearest measures yet of how generative AI has reshaped the web in under four years. Pew analyzed nearly half a million webpages drawn from the Common Crawl web archive, spanning January 2021 through July 2026, and ran them through an AI detection model created by Open Pangram.pewresearch
The Numbers
In a random sample of 10,000 webpages collected in July 2026, 10% showed "significant signs of AI authorship." But because the open web contains a mix of old and new material — much of it predating AI writing tools — Pew filtered out pages published before ChatGPT's launch. Among the remaining pages, 35% showed signs of AI authorship.techcrunch+1
The trend is not evenly distributed. Pages on .com domains show AI authorship at roughly 10 times the rate of .edu or .gov domains, which both hover around 1%. Pages on .org domains fall in between at 4.6%.pewresearch+1
Linguistic Tells
Pew identified specific linguistic markers that have surged across post-ChatGPT web content. Comparing recent snapshots to early 2023, em dashes now appear about twice as frequently, Oxford commas have increased 63%, and AI-favored vocabulary — words like "delve," "interplay," "tapestry," and "testament" — has more than doubled in usage. The use of "negative parallelism" structures ("it's not just X, it's Y") has nearly tripled.pewresearch
The study arrives shortly after Cloudflare reported that bot web traffic had overtaken human web traffic, a milestone reached sooner than the company had estimated.techcrunch
Caveats and Context
Pew acknowledged that AI detection tools can misclassify individual pages. "Models like this one look for statistical differences in how human authors and AI models use language," the report noted, adding that the patterns become more reliable at scale.pewresearch
The research was conducted by Samuel Bestvater, a senior data scientist at Pew, and contributes to a growing body of evidence suggesting the web is increasingly a place where, as TechCrunch put it, "much of it is bots reading webpages written by other bots."techcrunch
Sources (4)
- 1 How Much of the Internet Is Written With AI? www.pewresearch.org
- 2 A third of web pages published since ChatGPT's launch show signs of AI authorship, study finds | TechCrunch techcrunch.com
- 3 How many webpages are now written by AI? www.govtech.com
- 4 A third of webpages published since ChatGPT's launch show signs of AI authorship, study finds tech.yahoo.com
