다음을 통해 공유


Responsible AI for Microsoft Security Store

Microsoft Security Store is a unified destination for discovering Microsoft and partner security solutions, agents, and services that integrate with Microsoft Security products such as Microsoft Defender, Microsoft Sentinel, Microsoft Entra, Purview, Intune, and Security Copilot. As part of this experience, Microsoft is introducing AI-powered capabilities that help customers explore security outcomes and evaluate solutions more effectively.

This page provides a high-level overview of how Responsible AI (RAI) principles apply to the Security Store. Detailed information for specific AI features, such as Security Store Advisor is provided in dedicated feature-level responsible AI FAQ.


Microsoft’s approach to responsible AI

Microsoft designs and evaluates AI systems according to industry-leading standards for safety, privacy, and reliability. Our approach is anchored in six key principles:

  • Fairness

  • Reliability and safety

  • Privacy and security

  • Inclusiveness

  • Transparency

  • Accountability

Learn more: Microsoft Responsible AI principles


How AI is used in Security Store

Today, AI in the Security Store is used exclusively in the following feature:

Security Store Advisor

Security Store Advisor uses natural language processing and scenario-matching to help customers:

  • Discover Microsoft and partner security solutions for their security outcomes

  • Compare offerings using offer or solution metadata for objective comparison

Advisor operates entirely in unauthenticated mode and doesn't use tenant or user identity data.

For details about how AI works in this experience including intended use, evaluation, limitations, and privacy practices, see: Responsible AI FAQ for Security Store Advisor


Privacy and data handling

All Microsoft Security Store experiences, including AI-powered features, follow Microsoft’s standard privacy practices.

  • Security Store Advisor does not store user queries.

  • No tenant or identity data is accessed or used in recommendation generation.

  • All interactions are processed securely and transiently.

Learn more: Microsoft privacy