AI assistant message showing content violates uer guidelines

Forum|Forum|8 months ago
July 6, 2025
2 commentaire
304 vue

I am trying to get AI on Adobe Acrobat Reader to answer a question in a PDF document, but it says 'some content in the document question and or answer violates the user guidelines'. However, the PDF is a document that was a peer-reviewed academic paper published by the UK Government about child s**ual exploitation. It should not show this error message.

Adobe Applications

V

Vinod Dobhal

Adobe Employee

Hi @frosty_2085,

Thanks for flagging this issue. The model currently has safeguards that trigger when it detects content that could be considered explicit, whether it appears in the question you're asking or in the content of the response. These protections are designed to prevent misuse.

We'll take a closer look at this specific case to understand what triggered the message and see how we can improve the system's handling of serious, legitimate topics like this one.

Appreciate your patience, and thanks again for bringing it to our attention.

Regards,

Vinod

F

frosty_2085Auteur

Participant

Thanks. Yes, this is problematic for many people who need to use the tool to review academic papers that contain sensitive material, including those that have been peer-reviewed and therefore are not malicious in any way or intend to breach any guidelines.