An analysis of nearly 490,000 English-language webpages found that AI-associated writing has become a measurable part of the public web, growing steadily since generative AI tools became widely available.
More than one-third of pages published after ChatGPT became publicly available in November 2022 showed signs of AI involvement, following several years of growing access to generative AI tools.
AI-associated writing is appearing at different rates across the internet, with commercial websites showing higher rates than nonprofit, education and government domains.
Certain vocabulary and writing patterns have also become more frequent across webpages since 2023.
Why It Matters: AI is changing the makeup of the public web. As more online material includes AI-assisted writing, the information available to search engines, enterprise AI tools and agents increasingly includes a mix of human- and machine-produced content. For technology teams, understanding how that mix is changing can provide useful context for how systems retrieve, interpret and use information from the web.
- AI-Associated Writing Is Now a Fixture of the Web: The analysis examined webpages collected through the Common Crawl web archive between January 2021 and July 2026. In the July 2026 sample, 10% of all pages showed significant signs of AI writing or editing. The increase began around ChatGPT’s November 2022 release and has continued in the years since.
- Newer Pages Show How Far Adoption Has Moved: When the July 2026 sample was limited to pages published after ChatGPT’s launch, more than one-third showed significant signs of AI authorship or editing. Older material makes up a large portion of the internet, which helps explain the difference between the 10% rate across the full sample and the much higher rate among newer pages.
- Adoption Varies Across the Web: Signs of AI involvement appeared at similar rates across .com, .org, .edu and .gov domains around the time ChatGPT became available. By 2026, around 10% of .com pages showed signs of AI authorship, compared with 4.6% of .org pages and roughly 1% of .edu and .gov pages.
- Online Writing Is Changing Alongside the Tools: Several language features associated with AI models have become more frequent across webpages since 2023. Em dash usage has roughly doubled, Oxford comma usage increased 63% and AI-favored vocabulary more than doubled in use. Negative parallelism, including constructions similar to “it’s not just X, it’s Y,” nearly tripled, though it remains relatively uncommon.
- The Findings Measure Patterns Across the Web: Human writers use many of the same words and writing habits associated with AI-generated text, so none of these features can establish authorship on their own. Researchers used Open Pangram to identify statistical patterns in word choice and sentence structure across nearly 490,000 webpages. Detection models can misclassify individual documents, making the findings most useful for tracking how AI-associated writing has changed across large collections of online content over time.
Go Deeper -> How Much of the Internet Is Written With AI? – Pew Research Center
