AI-103: Developing AI Apps and Agents on Azure

By Microsoft · Question Mode

QuestionQ53

VerifiedImplement computer vision 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 image moderation to block unsafe content before the images are processed.

Does this solution meet the goal?

Explanation

Azure AI Content Safety image moderation screens visual content against harm categories such as violence, sexual content, self-harm, and hate, which addresses the unsafe-image half of the risk. It does not analyze any text extracted from the image for embedded instructions, so hidden prompt-injection content passed through OCR/text-extraction still reaches the model unmitigated, leaving the goal only partially met.

Learn more

Community Discussion

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