Research project: economic and infrastructural risks to digital archives posed by AI-driven content extraction

[Rasa Bočytė]

Dear all,

Colleagues at the University of Edinburgh are setting up a very timely research project related to AI-driven content extraction. Please read all the details below and reach out to Caitlin McDonald (caitlin....@ed.ac.uk) and Morgan Currie (morgan...@ed.ac.uk) if you would like to contribute.

--------------

Economic and infrastructural risks to digital archives posed by AI-driven content extraction

The University of Edinburgh is planning an academic research project tentatively titled “Economic and infrastructural risks to digital archives posed by AI-driven content extraction.” This will entail a series of interviews with key archival and commons-based data sources such as Wikimedia, the Flickr Foundation, the National Library of Scotland, the British Library, BBC Archive, and other large and small institutions impacted by these risks. (Note: we aim to interview people at these institutions but these specific examples are yet to be confirmed.) 

We need your help! We’re looking for additional contacts to interview across the EU. We’re particularly interested in hearing from archival bodies - whether public institutions, legal deposit libraries, charities, nonprofits, foundations, or other similar entities - which have either a legal mandate or a commons-based public interest in archiving any kind of content which is threatened by mass AI scraping tools. If that seems broad, it is! We think this is an issue impacting many organisations and we’re very curious to know what impacts they’re noticing from AI scraping and how they are mitigating these challenges or planning to do so in future.

Reach out to Caitlin McDonald (caitlin....@ed.ac.uk) and Morgan Currie (morgan...@ed.ac.uk) if you would like to participate in an interview and share your experience.

--

Kind regards,

Rasa Bocyte | she/her

Senior Advisor for Research Collaborations

Images of AI suck. Make better ones using the Archival Images of AI Playbook.

Media Parkboulevard 1, 1217 WE Hilversum
Postbus 1060, 1200 BB Hilversum | beeldengeluid.nl