Humanity's Memory: Internet Archive Takes Root in Switzerland

Thirty years ago, Brewster Kahle set out with the vision of enabling universal access to knowledge. What began in 1996 as an ambitious project by the US computer scientist with the Internet Archive and the now especially well-known Wayback Machine has become part of the digital backbone of collective memory. However, the internet is currently being radically transformed by generative AI, posing new hurdles for knowledge preservation. In this environment, the organization is now opening a new chapter with the founding of the Internet Archive Switzerland in St. Gallen.

The choice of location is no coincidence: with its UNESCO-protected Abbey Library, St. Gallen boasts over a thousand years of archiving tradition. According to the managing director of the Swiss branch, Roman Griesfelder, the academic environment there offers the ideal breeding ground to carry universal knowledge a step further into the future in a secure environment.

AI Models as New Cultural Heritage

According to the institution, the federal foundation operates legally independently but is part of a global network of independent libraries. A key focus is on securing endangered archives worldwide. A topic that will also be at the forefront of a UNESCO conference planned for Paris in November. In parallel, the foundation is dedicating itself to the Gen AI Archive in cooperation with the University of St. Gallen: under the direction of Professor Damian Borth, AI models are to be systematically archived for the first time.

This is a race against time. Websites are already considered ephemeral media. But AI models are changing so rapidly that their historical documentation overwhelms traditional archives. To understand in the future how algorithms shape today's society, the institution wants to preserve not only the results of AI but also the models themselves. Further Internet Archive branches already exist in and for Canada and Europe as a whole.

Collateral Damage in the AI War

In parallel, resistance against the efforts of online historians is growing. More and more publishers, such as The New York Times, are blocking the Wayback Machine's crawlers: they fear their content could be used as training material for AI giants like OpenAI without permission. The director of the search engine, Mark Graham, sees the archive thus becoming collateral damage in a conflict that is actually about copyright and royalties and should be resolved at that level.

It often turns out that even large media houses cannot maintain their own digital history without gaps. Graham reports of editors who found material missing from their own archives. This was then only findable in the Wayback Machine. Experts like media lawyer Kendra Albert warn: those who block archive bots across the board to keep AI scrapers out risk deleting the cultural memory of the present. The new Swiss presence is intended to serve as a stable anchor for public accessibility here.

(nen)