Questions

What they want to know is what type of files can it be used for. They are wondering if they could use it for open access published articles. From a copyright perspective, we are not necessarily concerned about the use of open access material, more the use of subscription material. Can you tell me if Adobe’s AI assistant copies the material to train the model? I will have a look at their terms and conditions, but if you could advise me on it, that would be most helpful.

Their published information about how the document content is used and handled is at https://helpx.adobe.com/acrobat/using/data-usage-and-handling.html

And asserts that it does not use the content to train the LLMs that they use.

However, I would say (and this is more DP area than ours) the difficulty always arises from the content of the document itself, so consideration that it doesn't fall into GDPR areas (i.e. there's personal information or it's identifiable, or it's used to make automated decisions about a person) should be considered by the user.

This product may be purchased by an individual, and this would typically place them outside the controls and constraints offered to the Unversity as an organisation.