I came across an interesting report from the GLAM-E Lab that raises a question we may not be discussing enough.
What happens when AI crawlers access the digital collections of museums, libraries, archives and galleries at enormous scale?
The report surveyed 43 cultural institutions. 39 reported increased website traffic, while 27 linked at least some of that increase to AI-training bots.
That may sound like a technical issue, but it has a very real consequence.
Heavy automated traffic can put pressure on servers, slow websites and, in some cases, force institutions to take parts of their collections offline.
There is an interesting tension here.
Cultural institutions are built around *public access*. Their purpose is to make knowledge, history and cultural material available to researchers, students and the wider public.
But when automated systems access that same material at machine scale, someone still has to pay for the infrastructure that keeps it online.
And simply using robots.txt may not be enough.
For me, the bigger question is not whether AI should access cultural heritage.
It is:
*Who should bear the cost of keeping the cultural web open when machines become its heaviest users?*
That conversation matters, because the unintended consequence of making collections freely available could be making them harder to access in the future.
*Source:* GLAM-E Lab www.glamelab.org/products/are-ai-bots-knocking-cultural-heritage-offline/