A third of webpages published since ChatGPT show signs of AI authorship
A third of English-language webpages published since the launch of ChatGPT in November 2022 show signs of being written or substantially edited by artificial intelligence, according to a Pew Research study released on Thursday.
Source: TechCrunch · August 20, 2026 at 5:27 PM · AI-assisted report
Single-source
KUALA LUMPUR, 21 AUGUST 2026 —
Listen to this article
DomainFork Audio · read aloud
AI-Generated Web Content Surges to One-Third of New Pages Post-ChatGPT, Study Finds
Market Impact
KUALA LUMPUR, Aug 21 — Over one-third of webpages published after ChatGPT’s launch in November 2022 show signs of being written or edited by artificial intelligence (AI), according to a study by Pew Research released on Thursday. The findings highlight the rapid integration of AI tools into digital content creation, reshaping the online information landscape.
The Pew Research report analyzed nearly half a million English-language webpages from the past five years using the Common Crawl web archive. Researchers employed Open Pangram’s detection technology to identify AI-generated or heavily edited content. In a random sample of 10,000 pages collected in July 2026, 10% showed significant signs of AI authorship. However, after filtering out older pages published before ChatGPT’s release, the share of AI-generated content rose to 35%.
The study also noted a tenfold higher rate of AI authorship in.com domains compared to.edu or.gov domains, which had AI authorship rates of around 1%..org domains recorded a 4.6% rate of AI-generated content.
The surge in AI-generated web content coincides with Cloudflare’s earlier report that bot web traffic had surpassed human traffic—a shift the company had not anticipated occurring so soon. While Cloudflare’s findings focus on web traffic patterns, Pew’s study examines the content itself, revealing that much of the newer web may be produced or modified by AI systems.
Domain-specific analysis showed stark contrasts in AI adoption. Commercial (.com) websites exhibited the highest incidence of AI-generated content, reflecting their reliance on scalable, automated content production. In contrast, educational (.edu) and government (.gov) domains maintained minimal AI involvement, likely due to stricter content oversight and regulatory requirements. Non-profit (.org) domains fell in between, with a 4.6% AI authorship rate.
The detection methodology, while at scale, is not without limitations. AI detection tools like Open Pangram can misclassify human-written content as AI-generated, and vice versa. However, Pew Research noted that the overall trends—such as increased use of em dashes, Oxford commas, and repetitive phrasing patterns—align with known characteristics of AI-generated text. These linguistic markers have become more prevalent in recent years, further supporting the study’s conclusions.
The implications for Malaysia’s digital ecosystem remain unclear, as the study does not provide country-specific data. However, the global trend suggests that AI-generated content is becoming a dominant force in online information dissemination. For Malaysian businesses, content creators, and policymakers, the findings underscore the need for greater awareness of AI’s role in shaping digital content and the potential challenges in distinguishing between human and machine-generated information.
Details not yet available on how Malaysian enterprises or regulatory bodies are responding to this trend.