Editorial illustration for AI made up over a third of new sites by 2025; Pope warning flagged as AI
AI-Generated Content Dominates Over One-Third of Web Sites
AI made up over a third of new sites by 2025; Pope warning flagged as AI
Why does a fake papal statement matter to anyone who builds a website? The episode began when a detection tool flagged a viral warning attributed to the Pope as AI‑generated, sparking a debate about how often machines masquerade as human voices online. Researchers at Stanford, Imperial College London and the Internet Archive have now turned that curiosity into a systematic audit.
By repurposing the earlier Pangram suite, they scanned every domain launched in the past two years, looking for the tell‑tale statistical fingerprints of machine‑written text. Their findings suggest that the problem isn’t limited to high‑profile hoaxes; it may be reshaping the very fabric of the web itself. As the line between authentic and synthetic content blurs, the scale of the issue becomes a crucial metric for policymakers, developers and anyone trying to trust what they read.
The numbers they uncovered are striking—and they set the stage for the claim that follows.
Text generated at least in part by AI accounts for more than a third of all new websites as of 2025, according to a study published this month by researchers at Stanford University, the Imperial College of London, and the Internet Archive. (The researchers used earlier Pangram tools to arrive at their findings.) It's this mess that Max Spero, CEO of Pangram and a self-professed "slop janitor," wants to help clean up. He tells WIRED that adding instant analysis to the company's browser extension offers people a more seamless way of checking for AI content across the sites they frequent.
The study’s headline figure—AI‑generated text making up more than a third of new sites by 2025—carries weight, given the involvement of Stanford, Imperial College and the Internet Archive. Yet the methodology, described only as using earlier Pangram tools, is not detailed in the excerpt, leaving open how the proportion was measured. Meanwhile, a detection tool flagged a recent Reddit post about the Pope’s warning as AI‑generated, even though the post itself was a routine personal‑care query on r/AmItheAsshole.
Could the tool be over‑sensitive, or does it reveal a broader trend of AI‑crafted content slipping into everyday discourse? The article notes the post was succinct, straightforward and grammatically clean, characteristics that often align with machine‑written prose. Still, without independent verification, the claim's still uncertain.
As AI continues to populate the web, distinguishing authentic human voices from algorithmic output will likely require clearer standards. Whether current detection methods can keep pace is, for now, an open question.
Further Reading
Related Reading
Research & Benchmarks Hyperparameter Tuning Reaches 0.9617 Accuracy in 64.59 Seconds
Research & Benchmarks Pharma Cautious as AI Promises Faster Drug Discovery and Smarter Trials
Research & Benchmarks Google AI Advisors Let Users Probe Performance with Conversational “Why” Queries
Same Week 91% of businesses now use video marketing — AI cut the cost of keeping up by 91% too
Market Trends
From Archives Leaks show pricing, specs of DJI’s upcoming beginner drones, not slated for US
From Archives Sergey Brin pushes DeepMind to match Claude, unveils agent skills catalog
Common Questions Answered
How much of new websites are estimated to be AI-generated by 2025?
According to a study by researchers from Stanford University, Imperial College London, and the Internet Archive, over one-third of new websites are expected to contain text generated at least partially by AI by 2025. This finding highlights the rapid integration of AI-generated content across the digital landscape.
Who is Max Spero and what is his role in addressing AI-generated content?
Max Spero is the CEO of Pangram and describes himself as a 'slop janitor' who aims to help clean up the proliferation of AI-generated content online. His company is working on tools and methods to analyze and potentially mitigate the spread of AI-generated text across websites.
What triggered the research into AI-generated website content?
The research was sparked by a viral statement attributed to the Pope that was flagged as AI-generated by a detection tool, which prompted researchers to investigate the broader phenomenon of AI masquerading as human-generated content. This curiosity led to a systematic audit of domains launched in the past two years using the Pangram suite of tools.