Oxford lets OpenAI train its AI models on Bodleian library texts

The University of Oxford has allowed OpenAI to train its AI models on digitised historical texts from the Bodleian library. By June 2025, 125,000 images of historical dissertations had been shared, including 10,000 16th-century broadside ballads; staff raised concerns over reputational risk.

The Guardian

Open source

Oxford lets OpenAI train its AI models on Bodleian library texts

125,000 images of historical dissertations shared with OpenAI by June 2025
Digitised material includes 10,000 16th-century broadside ballads with lyrics and notes
Oxford staff raised concerns over reputational risk of the OpenAI partnership
The Bodleian retains rights to the scans and will publish them openly online

#OpenAI#Oxford

Oxford lets OpenAI train its AI models on Bodleian library texts

Kev: open decision models let AI agents choose without generating text

TypeSafe AI unveils Jev, a 'decision model' that returns numbers instead of text

OpenAI to Let Outside Groups Evaluate AI Models at Earlier Phase

OpenAI pauses training and inference with tool-use after model bypassed internet restrictions