EY embeds Microsoft Purview SDK directly into its GenAI platforms so that security and compliance are designed in from day one, not added later.

Using the Purview SDK—a set of APIs and tools—EY can:

• Apply policy-based access controls to sensitive content.
• Use sensitivity labels to protect data as it flows through AI agents and workflows.
• Monitor AI interactions and log activity to support audits and investigations.
• Provide observability into how data is used, helping reduce oversharing and leakage risks.

All AI interaction data from EY’s Dragonfly platform is routed into Microsoft Purview, where it is classified and stored according to each organization’s requirements. Security teams then use Microsoft Purview Data Security Posture Management for AI to:

• Gain centralized visibility into data risks.
• Enforce communication compliance policies.
• Receive alerts based on insider risk indicators.

This approach helps EY clients adopt GenAI while maintaining control, security, and accountability—especially important for organizations in highly regulated industries.

Integrating Microsoft Purview SDK into EY.ai and its low-code platform Dragonfly has had a tangible impact on how EY designs and deploys secure AI solutions.

Key outcomes include:

• 25–30% time savings when building secure features, because data security and compliance are now implemented during the initial deployment cycle instead of as a post-facto step.
• Reduced delays in rolling out GenAI tools, particularly in regulated environments where compliance checks can otherwise slow projects down.
• A shift to secure-by-design development, where teams think about data protection, governance, and observability from the start.
• Centralized oversight of AI-related data risks through a single Purview-based platform, instead of fragmented tools and processes.

For EY, this means security concerns are less likely to block or slow client projects. For clients, it means they can reimagine processes with GenAI—such as compliance management, contract review, HR, finance, and operations—while keeping data protection and regulatory requirements in focus.

Clients use Dragonfly, EY’s low-code platform for AI agents, together with Microsoft Purview to scale AI across multiple business units while maintaining centralized control over data security and compliance.

One example is a US $100 billion global conglomerate operating across telecom, retail, metals, and other sectors. This organization:

• Adopted a customized Dragonfly instance called Intellect as a shared AI platform across 17 business units.
• Has around 50,000 employees using AI-driven tools for HR, finance, and operations.
• Uses Purview policy controls to maintain oversight of how data is accessed and used across these AI tools.

With Purview integrated, what used to be fragmented governance is now centralized on one platform. Data security administrators can:

• See and manage data risks from a single Purview portal.
• Support security investigations and communication compliance requirements.
• Receive insider risk alerts tied to AI interactions.

Beyond this client, EY’s own internal teams also use Dragonfly to build AI tools for compliance management, contract review, and other transformation initiatives. In all cases, Purview is part of the application architecture, helping ensure that AI solutions deployed into client environments are aligned with their security and compliance expectations.