One-Third of Web Pages are Now Written With AI, Says Pew Research
One-Third of Web Pages are Now Written With AI, Says Pew Research (pewresearch.org) 11
A Pew Research senior data scientist analyzed 500,000 English-language webpages from the past five years with an AI detection tool to estimate how many were likely written or edited by AI:
Of all the pages in this sample, 10% show significant signs of AI authorship... If we filter old webpages out of our samples and look at only the pages published after the release of ChatGPT, the trend is even more pronounced. In the July 2026 snapshot, signs of AI authorship can be found in over one-third of pages published after ChatGPT was released. This is in line with other studies that have shown that large shares of recently published pages on the internet were likely written or substantially edited by AI.
AI-authored text is not evenly spread across the web. When ChatGPT was first released, the kinds of linguistic patterns that can signal AI authorship appeared at similar rates across the main top-level web domains (.com, .org, .edu and .gov). But in samples from 2026, around one-in-ten pages with a .com domain show signs of AI authorship β about double the share on .org domains (4.6%), and 10 times the rate on .edu or .gov domains (both around 1%).
They even identified specific ways online text has changed, "comparing the internet of today to a snapshot from 2023."
Of all the pages in this sample, 10% show significant signs of AI authorship... If we filter old webpages out of our samples and look at only the pages published after the release of ChatGPT, the trend is even more pronounced. In the July 2026 snapshot, signs of AI authorship can be found in over one-third of pages published after ChatGPT was released. This is in line with other studies that have shown that large shares of recently published pages on the internet were likely written or substantially edited by AI.
AI-authored text is not evenly spread across the web. When ChatGPT was first released, the kinds of linguistic patterns that can signal AI authorship appeared at similar rates across the main top-level web domains (.com, .org, .edu and .gov). But in samples from 2026, around one-in-ten pages with a .com domain show signs of AI authorship β about double the share on .org domains (4.6%), and 10 times the rate on .edu or .gov domains (both around 1%).
They even identified specific ways online text has changed, "comparing the internet of today to a snapshot from 2023."
- Em dashes appear twice as often
- Oxford commas increased 63%.
- Words used more frequently by AI (like "delve," "interplay" or "testament") have more than doubled in usage.
- "Negative parallelism" comparisons ("it's not just X, it's Y") have nearly tripled.
Re: no surprise (Score:2)
And the last 15 years the corporate web pages have been dumbed down to irrelevancy so you can't find any useful info anymore.
I think it's missing some nuance (Score:3)
It's not the interplay between the decrease in literacy [futurism.com] in high school and college and artificial intelligence's improvement in text generation, it's a testament to the complexity of the problem β being critical to delve into the cognitive, cultural, and other foundations of human language and thought β in addition to examining not just the production, but also the consumption of language by both the human brain and large language models.
Re: (Score:1)
β
I see what you did there.
Re: I think it's missing some nuance (Score:2)
Breaking news (Score:2)
A new tool that's really good at written language is being used by folks to edit copy. And the unsavory are using it to externalize costs and monetize slop. All this seems to distract everyone from the big issues like ownership and mass surveillance.
Expect "make it look human" post-processing (Score:1)
Expect AI-written material to be put through a "cleanup to make it look less like AI and more like a person" post-processing step in the very near future.
It's probably already happening.
Mass spamming (Score:2)
Yeah there have been tons of spam site popping up that are now produced with AI text instead of less sophisticated templates, it would be interesting to see if there's a downtick in the previous spam site markers.
SEO bait pages that we didn't need anyway (Score:2)
Google Search results are a joke these days.
Maybe Firefox should build a filter deleting all such pages.
Easy fix (Score:2)
Confounding variables (Score:2)
I'm curious how they dealt with the fact the language changes over time. It makes sense that people writing after AI would write more like an AI than people from before AI.
How it works
Once you click Generate, Ollama reads this article and crafts 5 comprehension questions. Your answers are graded against the article content β general knowledge won't be enough. Score 70+ to count toward your certificate.
Questions are cached β you'll always get the same 5 for this article.