Microsoft Purview | Communication Compliance: Detect potentially risky generative AI interactions

Microsoft Purview's Communication Compliance now offers a new feature that detects potentially risky generative AI interactions. This new feature includes classifiers for prompt injection attacks and protected material, powered by Microsoft Azure AI Content Safety. It helps identify risks of prompt injection attacks and can identify when generative AI responses contain branded or copyrighted material.

The Public Preview for this new feature was rolled out in mid-November 2024, and General Availability details will follow. Communication Compliance admins do not need to take any admin action before the rollout. The rollout will happen automatically by the specified date.

Once the new classifiers are visible in an organization's tenant, they can be configured in any Communication Compliance policy that focuses on a generative AI workload. Learn more about the new feature in the Trainable classifiers section of Create and manage communication compliance policies in Microsoft Learn.

This new feature empowers organizations to maintain their content's originality, protect their reputations, and identify potentially malicious generative AI interactions.

The post Microsoft Purview | Communication Compliance: Detect potentially risky generative AI interactions originally appeared on M365 Admin.