An analysis of news summaries reveals that since 2022, approximately 35 percent of newly created website content has been generated by artificial intelligence. Researchers from Stanford University, Imperial College London, and the Internet Archive found that AI has transformed online writing styles into a more positive, concise, and effective format. The study indicates that AI has not increased the spread of false information but has reduced linguistic diversity, pushing the internet towards uniformity.
Following the public release of ChatGPT in late 2022, there has been a rapid transformation of the digital landscape. The research highlights that AI has enhanced online writing by making it more positive, polite, and capable of conveying much in fewer words. The study titled “The Impact of AI-Generated Text on the Internet” challenges the belief that AI is responsible for spreading misinformation online.
Tests conducted by fact-checkers found that AI did not cause a significant rise in false claims; however, there was a noticeable decrease in linguistic and stylistic diversity. Since many AI models tend to write in a similar, standardized style, human creativity and unique voices are diminishing, raising concerns that the internet may become increasingly homogenized, according to the study.
The research team collected website samples from August 2022 to May 2025 using the Wayback Machine and analyzed them with the assistance of the “Pangram v3” software. Researchers emphasize that the next major challenge is to develop AI as a partner in human creativity rather than allowing the internet to become solely uniform and machine-generated. Although based on results from a specific timeframe, the study outlines plans to continually monitor these trends in collaboration with the Internet Archive, developing tools to track ongoing changes.
