One-Third of Post-ChatGPT Web Content Is AI-Written, Pew Study Finds
A scan of 500,000 webpages reveals a surge in automated content, particularly across commercial domains.
The landscape of the public internet has shifted fundamentally since the debut of generative AI, with a significant portion of new web content now produced by machines. A recent study by Pew Research indicates that approximately one-third of the web published after the release of ChatGPT is AI-written.
To reach this conclusion, Pew Research conducted a large-scale scan of nearly 500,000 English-language webpages sourced from the Common Crawl archive. Using AI detection tools, researchers found that roughly 35% of these pages show signs of AI authorship or substantial automated editing. The data reveals a stark divide in how different sectors of the web are adopting the technology; AI-written content is heavily concentrated on .com domains, appearing at roughly 10 times the rate of .edu or .gov domains.
The Rise of Automated Publishing
This surge follows the widespread release of ChatGPT and other large language models, which granted users the ability to generate high-volume, human-like text almost instantaneously. This capability has fueled a proliferation of automated content, often designed to satisfy search engine optimization (SEO) algorithms rather than provide unique human value. The trend has reignited discussions around the "dead internet theory," the hypothesis that the majority of internet traffic and content is now generated by bots rather than humans.
Implications for the Digital Ecosystem
The concentration of AI content on commercial (.com) domains suggests a systemic shift in how information is produced for profit. As the web becomes saturated with synthetic text, the reliability of web-based information comes into question. There is also a growing concern regarding a "model collapse" feedback loop, where future AI models are trained on the very AI-generated data currently flooding the internet, potentially degrading the quality and accuracy of future iterations.
The Future of Search
As AI-generated content continues to multiply, the future of search engine optimization remains uncertain. Search engines must now evolve to distinguish between high-value human expertise and high-volume synthetic output. While the Pew study provides a snapshot of the current state of the web, it remains to be seen how the continued evolution of detection tools and search algorithms will impact the visibility of AI-authored content in the coming years.