Microsoft Identifies and Mitigates AI-Driven Data Risks in Fabric
Microsoft Enhances Data Security and Governance in Fabric with AI-Driven Innovations
Microsoft has introduced new features in Microsoft Purview to help organizations secure their data and accelerate the adoption of artificial intelligence (AI) in Microsoft Fabric. The updates focus on identifying and mitigating data risks, ensuring the protection of sensitive data, and maintaining visibility across the data environment.
Unified Foundation for Data Security and Governance
The integration of Microsoft Purview and Microsoft Fabric provides a unified foundation for data security and governance, enabling organizations to protect sensitive data and maintain data readiness for AI use.
“This integration enables organizations to innovate with confidence, ensuring data is protected, governed, and trusted for responsible AI activation.” – Darren Portillo, Product Marketing Manager at Microsoft
Data Protection and Governance Capabilities
Microsoft Purview provides data protection capabilities across the Fabric data estate, including information protection and data loss prevention (DLP) policies. These policies help restrict access to sensitive data and detect potential data exfiltration. Insider risk policies for Fabric also extend to Fabric lakehouses, providing built-in risk indicators based on user activity.
AI Data Exposure Management
To manage AI data exposure, new preview capabilities support risk detection in Copilot and agent interactions. Organizations can identify sensitive data in prompts and responses and take action to reduce exposure. Additionally, data risk assessments can be conducted to identify sensitive or unprotected Fabric assets, enabling organizations to apply labels or policies to reduce exposure.
Audit, eDiscovery, and Retention
Microsoft Purview also supports governance of AI usage through capabilities such as audit, eDiscovery, retention, and detection of non-compliant activity. The Unified Catalog provides a centralized view of data assets to support enterprise data governance, enabling organizations to discover and use data while maintaining governance standards for AI scenarios.
Data Quality and Governance
Built-in data quality capabilities allow organizations to measure, monitor, and improve data quality, supporting trusted data for decision-making and AI. For Fabric OneLake catalog users, Microsoft Purview extends governance capabilities across data consumers and governance teams.
Controlled Data Publishing Workflows
Controlled data publishing workflows in the Unified Catalog enable data owners to manage the publication of data products and glossary terms, supporting consistent governance and maintaining a curated catalog. Quality checks can also be run on unmanaged data assets without linking them to data products, enabling faster validation at scale and supporting broader access to reliable data for AI use cases.
About Author
English