A Pew Research Center study published on August 20, 2026, reveals that 35% of English-language web pages published since the November 2022 launch of ChatGPT show significant signs of AI authorship. Analyzing 490,000 pages via the Common Crawl archive and the Open Pangram detector, researchers found that .com domains exhibit AI signs at ten times the rate of .edu and .gov sites. The rise of AI text is marked by a doubling of em dashes and specific vocabulary words like 'delve' since 2023.
AI authorship prevalence on web
- ▪A Pew Research Center study published on August 20, 2026, found that 35% of English-language web pages published after the November 2022 launch of ChatGPT show significant signs of AI authorship.
- ▪In a random sample of 10,000 English-language web pages collected in July 2026, approximately 10% showed significant signs of AI authorship.
Domain-level AI content distribution
- ▪The AI-authorship rate for .com domains climbed from approximately 1% in January 2021 to 9.35% in January 2026.
- ▪Web pages with a .org domain showed a 4.6% rate of AI authorship in the August 2026 Pew Research Center study.
- ▪Web pages with a .com domain show signs of AI authorship at approximately 10 times the rate of .edu or .gov domains, which both sit near 1%.
AI detection methodology
- ▪Pew Research Center analyzed approximately 490,000 English-language web pages from the Common Crawl web archive spanning January 2021 through July 2026.
- ▪Pew Research Center noted that AI detection models are not perfect and can misclassify individual human-written pages as AI-authored and vice versa.
- ▪Pew Research Center utilized Open Pangram, an open-weight AI detection model from Pangram Labs, to analyze the collected web pages for statistical patterns of AI authorship.
Linguistic markers of AI text
- ▪The use of 'negative parallelism' comparisons, such as 'it's not just X, it's Y,' has nearly tripled on web pages since 2023.
- ▪Since 2023, the use of em dashes has doubled and the use of Oxford commas has increased by 63% on web pages published after the launch of ChatGPT.
- ▪AI-favored words such as 'delve,' 'interplay,' and 'testament' have more than doubled in usage on web pages since 2023.
AI content ecosystem implications
- ▪Internet infrastructure provider Cloudflare reported that bot web traffic has overtaken human web traffic sooner than the company had previously estimated.
- ▪An analysis by Graphite found that while AI-generated articles approached half of newly published online articles, human-written articles still dominate Google search results and AI citations.
Story comments
Loading comments…