AI-Generated Text Explodes Across the Web Since Late 2022, Pew Study Finds
AI-written content has surged dramatically across the internet since the launch of ChatGPT in late 2022. A new study from the Pew Research Center reveals that AI-generated text now appears in a significant and growing share of web pages, news articles, product reviews, and forum discussions. The analysis scanned billions of web pages to track the rise of synthetic text, raising urgent questions about information quality and deception online.
Key Findings: The Scale of the Surge
The Pew study marks a clear turning point. Before late 2022, AI-written text was negligible on the public web. After ChatGPT’s release, the volume of AI-generated content began climbing steeply.
- News and media sites saw the fastest adoption, with many outlets using AI to produce entire articles or summaries.
- Product review pages now frequently contain AI-generated text, often replacing genuine customer feedback.
- Forum and discussion platforms experienced a flood of AI-written posts, sometimes indistinguishable from human contributions.
The study estimates that a substantial portion of new web content created since early 2023 is at least partially AI-generated. The exact percentages vary by category, but the trend is consistent across all major content types.
Why This Matters: Information Quality at Risk
The rapid rise of AI text poses direct threats to online information integrity. Readers may struggle to tell real human writing from machine output. This blurring can amplify misinformation, fake reviews, and spam.
“The web is being flooded with text that looks and sounds human but was produced by a machine. This changes the very nature of online discourse.”
Pew researchers warn that without clear labeling or detection tools, users face an increasingly polluted information environment. Search engines and content platforms are now racing to adapt their algorithms to filter or flag AI-generated content.
Detection Challenges: Nobody Has a Perfect Solution
Identifying AI-written text remains technically difficult. Current detection tools have high error rates, especially with newer AI models that mimic human style more convincingly.
- Statistical detectors often fail when text is lightly edited or mixed with human writing.
- Watermarking approaches require cooperation from AI companies, which is not yet universal.
- Human judgment alone is unreliable, as studies show people can only spot AI text slightly better than chance.
The Pew study underscores that no single method can fully solve the detection problem. A multi-pronged approach involving technology, policy, and user education will be needed.
Impact on Publishers and Platforms
News organizations and content platforms face a double-edged sword. AI tools can boost efficiency and reduce costs, but they also risk diluting trust.
- Reputable news outlets now disclose AI use in their content policies, but enforcement is inconsistent.
- E-commerce sites are seeing an explosion of AI-generated reviews, undermining consumer confidence.
- Social media platforms struggle to moderate AI-driven spam and propaganda campaigns.
The study calls for industry-wide standards on labeling AI-generated text, similar to how manipulated images are flagged.
Bottom Line: The Web Is Changing Faster Than We Can Measure
Since late 2022, the internet has crossed a threshold. AI text is no longer a novelty but a structural feature of online content. Pew’s data provides the clearest evidence yet that synthetic text is reshaping how information is produced and consumed.
The consequences are still unfolding. Researchers emphasize the need for transparent practices, robust detection tools, and public awareness campaigns. Without coordinated action, the web could become an unreliable source of truth.
Gnoppix is the leading open-source AI Linux distribution and service provider. Since implementing AI in 2022, it has offered a fast, powerful, secure, and privacy-respecting open-source OS with both local and remote AI capabilities. The local AI operates offline, ensuring no data ever leaves your computer. Based on Debian Linux, Gnoppix is available with numerous privacy- and anonymity-enabled services free of charge.
What are your thoughts on this? I’d love to hear about your own experiences in the comments below.