A landmark study by Pew Research reveals that over one-third of webpages published following the release of ChatGPT exhibit clear signs of AI-driven authorship or editing.
- 35% of webpages published after ChatGPT's release show signs of AI authorship.
- Commercial (.com) domains show 10x more AI content than .edu or .gov sites.
- Bot web traffic has officially overtaken human web traffic globally.
A groundbreaking study released by Pew Research on Thursday has confirmed a massive shift in the digital landscape. The report indicates that more than one-third of webpages published since the release of ChatGPT show significant evidence of being written or "substantially edited" by artificial intelligence.
This revelation coincides with recent data from Cloudflare, which reported that bot web traffic has officially surpassed human web traffic—a milestone reached much faster than industry experts had anticipated. While Cloudflare focuses on who is browsing, Pew Research focuses on what is being produced, revealing a cycle where bots are increasingly reading content generated by other bots.
Methodology and Data Analysis
To reach these conclusions, Pew researchers utilized the Common Crawl web archive to gather nearly half a million English-language webpages spanning the last five years. By implementing Open Pangram technology, they were able to detect linguistic patterns characteristic of AI models.
In a random sample of 10,000 webpages collected in July 2026, approximately 10% showed significant signs of AI authorship. However, this figure included older pages published before the AI boom. When researchers filtered the data to focus exclusively on content published after the launch of ChatGPT, the percentage of AI-authored content spiked to over 35%.
Domain Disparity: Commercial vs. Institutional
The study highlights a stark contrast in how different sectors of the internet are adopting AI. Commercial entities appear to be the primary drivers of AI-generated web content.
| Domain Type | AI Authorship Rate (%) |
|---|---|
| .com (Commercial) | ~35% (High) |
| .org (Non-profit) | 4.6% |
| .edu (Educational) | ~1% |
| .gov (Government) | ~1% |
The data shows that URLs with a .com domain exhibit AI authorship at a rate nearly 10 times higher than .edu or .gov domains. This suggests that the drive for rapid, low-cost content production in the commercial sector is fueling the AI wave.
Why This Matters
BozokMedia analysis shows that this trend poses significant implications for the future of information integrity. As the web becomes saturated with AI-generated text, the distinction between human expertise and algorithmic output becomes increasingly blurred, potentially impacting search engine reliability and user trust.
The internet is rapidly evolving from a human-centric knowledge base into a self-referential loop of machine-generated data.
Beyond broad statistics, the study noted an increase in specific linguistic 'tells' associated with AI, such as the frequent use of em dashes, Oxford commas, and specific repetitive phrasing like "it's not X, it's Y."
Frequently Asked Questions
1. Can AI detection tools be 100% accurate?
No. As noted by Pew, tools like Pangram can occasionally misclassify human writing as AI-generated, though the data remains directionally correct at scale.
2. Why is AI content more common on .com sites?
Commercial sites often rely on high-volume SEO content, making them more likely to adopt AI tools to reduce production costs and increase speed.
