EY uses Microsoft Purview SDK as a core part of its GenAI application architecture so that security and compliance are built in from the start rather than added later. Instead of creating custom safeguards from scratch, EY integrates Purview SDK’s policy-based controls directly into its GenAI platform, EY.ai, and into Dragonfly, EY’s low-code environment for building intelligent agents and AI workflows. Through this integration, EY can: - Apply sensitivity labels and protection policies to data used by AI agents. - Enforce policy-based access and protection controls on sensitive content. - Monitor AI interactions and log activity to support compliance and investigations. - Centralize visibility of data risks in the Microsoft Purview portal. This approach helps EY’s clients, many of whom operate in highly regulated industries, adopt GenAI while maintaining control over data security, compliance, and governance. Security and observability are designed into the platform, so AI tools can be deployed more confidently in complex, regulated environments.

EY reports clear, measurable benefits from using Microsoft Purview SDK in its AI development process. Key outcomes include: - **25–30% time savings on secure feature development**: Data security and compliance capabilities that previously took about an extra week to bolt on after initial development are now built during the first deployment cycle. EY estimates a 25–30% reduction in time when building secure features using Purview SDK. - **Faster rollout of GenAI tools in regulated environments**: Because security and compliance are embedded early, teams experience fewer delays related to risk reviews and remediation, which helps accelerate deployment timelines. - **Centralized oversight instead of fragmented controls**: By routing AI interaction data from Dragonfly into Purview, EY’s clients gain a single view of data classification, policy enforcement, and risk indicators, simplifying governance and reducing operational friction. Overall, Purview SDK helps EY move from reactive security add-ons to a proactive, secure-by-design model that shortens development cycles while maintaining compliance expectations.

Dragonfly is EY’s low-code platform for creating and managing AI agents and workflows on top of its EY.ai GenAI foundation. When combined with Microsoft Purview SDK, it gives large enterprises a way to scale AI while keeping data security and compliance in focus. Here’s how it works in practice: - **Integrated security and compliance**: Dragonfly is built with Purview SDK integrated upfront, so every AI agent and workflow can inherit policy-based access controls, sensitivity labels, and protection rules. - **Data routing and classification**: EY has created a pipeline that routes AI interaction data from Dragonfly into Microsoft Purview. There, the data is classified and stored according to each organization’s requirements. - **Centralized risk visibility**: Security and compliance teams use Microsoft Purview Data Security Posture Management for AI to see how AI agents are interacting with sensitive data, review logs, and respond to alerts, including insider risk indicators. - **Support for large, distributed organizations**: One example is a US$100 billion global conglomerate with 17 business units across telecom, retail, metals, and other sectors. The company uses a customized Dragonfly instance called Intellect as a shared AI platform for about 50,000 employees across HR, finance, and operations. Purview’s policy controls help maintain oversight across this diverse environment. - **Internal and client use cases**: EY’s own teams use Dragonfly to build AI tools for compliance management, contract review, and other transformation initiatives that must meet strict security and compliance standards before entering client environments. By combining Dragonfly with Purview SDK, EY helps enterprises reimagine how they adopt AI—moving from siloed, ad hoc tools to a shared, policy-driven platform where trust, security, and transparency are part of the design from day one.