Azure AI Content Safety vs Azure OpenAI content Safety

Question

Azure AI Content Safety vs Azure OpenAI content Safety

Jason Isaac 25

Hello All,

When working with AI services in Azure, we noticed there is a separate AI content safety service and a Default Content Safety that gets applied when deploying an Azure OpenAI model.

User's image

Are these 2 things same or different services?. if they are different services, can someone please provide more details on the differences.

When to use the default service and when to go for the separate content safety service?.

Thanks.

Accepted answer

0 additional answers

Your answer

Answer 1

Saideep Anchuri 9,500 Moderator

Hi Jason Isaac

Welcome to Microsoft Q&A Forum, thank you for posting your query here!

Azure AI Content Safety: This is a specialized service that helps find and manage harmful content created by users or AI in apps and services. It offers various tools to check text, images, and mixed content for safety, and has features like checking if content is accurate, protecting sensitive material, and adding extra safety layers.

Azure OpenAI Model Content Safety: This is a built-in safety feature for Azure OpenAI models that makes sure their outputs are safe and follow certain standards. It might not have as many detailed options or custom settings as the dedicated Azure AI Content Safety service.

kindly refer below link: What is Azure AI Content Safety

understand-content-filter

Hope this helps. Do let us know if you any further queries.

If this answers your query, do click Accept Answer and Yes for was this answer helpful.

Thank You.

Jason Isaac 25 Reputation points

2025-01-30T06:32:48.26+00:00

Hello @saideep anchuri ,

Thanks for your quick reply and explanation.

If I understand you correctly, they are different and the models used to detect the harmful content are different.

Could you please confirm my understanding?.

The reason I am asking this is because the harmful categories that are getting detected are same in both cases(Jailbreak, Hate, Sexual, Self-Harm, Violence) Azure AI content safety,openai content filters, and when deploying an Azure OpenAI model with default content safety these categories will be detected and it seems like I don't need a separate AI content safety service.

But as they are different, I might want to use a separate AI content safety service on top of the default setting.

Thank you.
Saideep Anchuri 9,500 Reputation points Moderator

2025-01-30T07:02:17.98+00:00

Hi Jason Isaac

Yes, you can add extra azure ai content safety resource for advance configuration like groundless detection, protected material detection etc. before sending it to open ai as the responses will be blocked or filtered with default content filter in open ai.

Kindly refer below link: content-filters

Thank You.
Jason Isaac 25 Reputation points

2025-01-30T07:08:39.17+00:00

Thanks for your support @saideep anchuri . Accepted your answer.

Share via

Azure AI Content Safety vs Azure OpenAI content Safety

0 additional answers

Your answer