You have a chatbot built on Azure OpenAI (AI1). The requirement is that all input and output from the chatbot must be checked for objectionable content (such as hate speech, sexual, violent, or self-harm content).
To meet this requirement, Microsoft recommends integrating Azure AI Content Safety with Azure OpenAI.
Azure AI Content Safety provides APIs to detect and filter harmful or objectionable text and images.
For chatbot scenarios, you can configure Content Safety as a moderation layer on both user input (prompt) and model output (response).
This ensures safe interactions before messages are returned to users.
Other options:
A). Azure Machine Learning - used for building, training, and deploying custom ML models, not for content moderation.
B). Log Analytics - used for monitoring and logging, not for filtering objectionable content.
D). Microsoft Defender Threat Intelligence (Defender TI) - provides cybersecurity intelligence, not responsible for moderating chatbot content.
Therefore, the correct resource to create is Azure AI Content Safety.
The answer: C. Azure AI Content Safety
Microsoft References
Azure OpenAI and Azure AI Content Safety integration
Azure AI Content Safety overview
Moderate text with Azure AI Content Safety