started · updated
Pew Research finds AI authorship in 10% of English webpages
A study by the Pew Research Center reveals that approximately 10% of English-language webpages sampled in July 2026 show signs of AI authorship or editing. This figure rises significantly to over 35% when analyzing only content published after the launch of ChatGPT in November 2022.
Using the Open Pangram detection tool to analyze nearly 500,000 pages from the Common Crawl archive, researchers found that AI usage is most prevalent in commercial domains. Specifically, .com domains show a 9.35% detection rate, while .org domains are at 4.59%. In contrast, academic (.edu) and government (.gov) sites remain much lower, at 1.03% and 0.76% respectively.
The research also highlights shifts in linguistic patterns. Since 2023, the use of em dashes has doubled, and the frequency of Oxford commas has increased by 63%. Additionally, the use of specific terms like 'delve', 'interplay', and 'testament' has more than doubled, while the rhetorical pattern 'it's not just X, it's Y' has increased approximately three times.
Entities
ChatGPT · Common Crawl · Open Pangram · OpenAI · Pew Research Center · Samuel Bestvater
Claims
What the coverage asserts, and how many sources carry each claim.
- [● 3 SOURCES] Approximately 10% of English-language webpages sampled in July 2026 show signs of AI authorship or editing. www.tigrishaber.com · www.businesstechweekly.com · www.panorama.it
- [● 2 SOURCES] The frequency of specific words such as 'delve', 'interplay', and 'testament' has more than doubled. www.tigrishaber.com · www.panorama.it
- [● 2 SOURCES] The rhetorical pattern 'it's not just X, it's Y' has increased approximately three times. www.tigrishaber.com · www.panorama.it
- [● 3 SOURCES] The study analyzed nearly 500,000 English-language webpages from the Common Crawl archive using the Open Pangram detection tool. www.tigrishaber.com · www.businesstechweekly.com · www.panorama.it
- [● 2 SOURCES] AI detection rates for .org, .edu, and .gov domains are 4.59%, 1.03%, and 0.76% respectively. www.businesstechweekly.com · www.tigrishaber.com
- [● 3 SOURCES] Over 35% of webpages published after the launch of ChatGPT in November 2022 show signs of AI authorship or heavy editorial contribution. www.tigrishaber.com · www.businesstechweekly.com · www.panorama.it
- [● 2 SOURCES] Since 2023, the use of em dashes has doubled and Oxford comma usage has increased by 63% on the web. www.tigrishaber.com · www.panorama.it
- [● 2 SOURCES] The AI detection rate for .com domains is 9.35%. www.businesstechweekly.com · www.tigrishaber.com