Pew Finds 35% of Post-ChatGPT Webpages Show AI Authorship
Updated
Updated · TechCrunch · Aug 20
Pew Finds 35% of Post-ChatGPT Webpages Show AI Authorship
2 articles · Updated · TechCrunch · Aug 20
Summary
Pew Research found 35% of webpages published after ChatGPT’s November 2022 launch show signs of AI authorship or heavy AI editing.
Nearly 500,000 English-language pages from Common Crawl were analyzed, and a July 2026 random sample showed only 10% AI authorship until older pre-ChatGPT pages were removed.
Domain patterns were uneven: .com pages showed AI authorship at roughly 10 times the rate of .edu or .gov pages, which were each near 1%, while .org stood at 4.6%.
Pew said detector errors can misclassify individual pages, but argued the large-scale trend is directionally reliable and aligns with rising AI-style markers such as em dashes and Oxford commas.
The report adds to evidence that AI is reshaping the web’s content layer, arriving soon after Cloudflare said bot traffic had already surpassed human traffic online.