On October 19, 2026, Brewster Kahle, founder of the Internet Archive, warned that publishers are restricting web archiving to prevent AI firms from exploiting content, inadvertently threatening the preservation of internet information history.
The Internet Archive was established in 1996 and developed the Wayback Machine to preserve website versions over time for research, journalism, history, and information verification.
The New York Times, The Guardian, USA Today, and many other news agencies have implemented archiving restrictions. By May 2026, over 340 local news sites in the US had limited access for the Internet Archive.
Publishers fear the Wayback Machine creates an access pathway bypassing paywalls and licensing mechanisms,enabling AI companies to collect unauthorized training data.
Kahle argued that blocking the non-profit library does not resolve AI companies’ data scraping activities, but diminishes resources for research, investigative journalism, and accountability.
A 2024 Pew Research Center study of 1 million web pages found that 25% were no longer accessible after 10 years; for pages dating back to 2013, the disappearance rate reached 38%.
The Internet Archive noted that the Wayback Machine preserved 16% of the disappeared URLs in the study’s dataset,highlighting the role of independent archives.
The Wayback Machine also preserves materials from defunct news agencies such as Rocky Mountain News, MTV News, Apple Daily, and Stand News.
The Internet Archive proposes distinguishing public-interest archiving from commercial AI data collection through access limits, traffic monitoring, and bulk download controls.
📌 Conclusion: The copyright dispute between publishers and AI companies is producing unintended consequences for the preservation of digital heritage. While 25% of websites surveyed by Pew vanished after 10 years, restricting the Wayback Machine risks widening the historical documentation gap. Brewster Kahle proposes separating public-interest content archiving rights from commercial data scraping rights, combining copyright protection mechanisms and access controls to maintain society’s digital memory.
