AI-103: Developing AI Apps and Agents on Azure

By Microsoft · Question Mode

QuestionQ54

VerifiedImplement generative AI and agentic solutions

You have a multimodal generative AI model that accepts image uploads and uses extracted image text to generate responses.

You discover that users can upload unsafe images and embed hidden instructions within images to manipulate the model.

You need to implement controls to mitigate this risk.

Proposed solution: Configure a prompt shield for documents.

Does this solution meet the goal?

Explanation

Prompt Shields for Documents inspects content the model did not receive directly from the user — such as text extracted from an uploaded image — for embedded, indirect prompt-injection instructions, which addresses the hidden-instruction half of the risk. It performs no visual content moderation, however, so it does not detect or block unsafe (violent, explicit, or otherwise inappropriate) image content, leaving that half of the stated risk unaddressed.

Learn more

Community Discussion

No comments yet. Be the first to start the discussion!