Roughly one in ten webpages sampled by Pew Research Center show detectable signs of AI authorship. Among pages published after ChatGPT's late 2022 debut, that share jumps to 35%. The numbers offer the clearest public estimate yet of how much of the open web is now machine-written, at least in part.
Pew's researchers used detection tools to scan a broad sample of English-language webpages, looking for linguistic patterns associated with large language model output. They then compared results against publication dates to isolate content created before and after ChatGPT went mainstream. The gap between the two eras — a jump from a low baseline to over a third of newer pages — is the headline finding, not the 10% figure itself.
It's worth being clear about what this study can and cannot tell us. AI detection tools are notoriously imperfect. They can flag heavily-edited human writing as machine-generated and miss AI text that's been lightly rewritten by a person. Pew's numbers are best read as a directional signal — the web is absorbing more AI-assisted content over time — rather than a precise headcount.
This isn't the first attempt to measure the phenomenon. Content-detection companies and academic researchers have published similar estimates over the past two years, with figures ranging from under 5% to well over 50% depending on methodology and what counts as "AI-authored." Pew's study adds institutional weight to a trend that marketers and SEO professionals have already been discussing anecdotally for a while.
The pattern fits a broader arc: as generative AI tools became free or cheap to access, content production costs for a certain kind of formulaic writing — product descriptions, roundup articles, basic explainers — dropped close to zero. Search engines and platforms have responded unevenly, penalizing obviously low-quality mass-produced content while stopping short of banning AI assistance outright. Google, for instance, has repeatedly said it rewards helpful content regardless of how it was produced, while also rolling out algorithm updates aimed at scaled, low-value AI content specifically.
For small businesses, this matters less as a curiosity and more as a competitive and trust issue. If a third of new web content is AI-touched, the bar for what makes your content stand out has likely moved. Generic, fast-produced text is now table stakes, not a differentiator — anyone can generate it in seconds, so it's less likely to earn rankings, backlinks, or reader trust on its own.
There's also a practical SEO risk. If search engines tighten enforcement against AI-generated content that's thin or unedited, businesses that have leaned heavily on AI drafting tools without human review or added expertise could see traffic dips with little warning. The safer posture, based on how search algorithm updates have played out over the past two years, is treating AI tools as a drafting aid rather than a finishing tool — human editing, original data, and firsthand experience still tend to separate content that performs from content that doesn't.
Finally, there's a reputational dimension. Customers and potential clients are increasingly primed to wonder whether a blog post, review response, or About page was written by a person. Being transparent about how you use AI tools — and where a human is still doing the thinking — may become a small trust signal in itself, not unlike disclosing sourcing in journalism.
Watch for two things over the coming months: whether major search engines announce new detection or ranking measures specifically targeting AI-authored content, and whether AI detection tools themselves improve enough to be used reliably in enforcement rather than just research. Right now, neither is settled.
The practical takeaway for small business owners: audit your own published content for anything that reads as generic or unedited AI output, and prioritize adding original insight, data, or experience to anything that's public-facing. The web is getting more crowded with AI text — differentiation, not detection-dodging, is the more durable strategy.