Microsoft to Retire Auto-Generated Semantic Models in Fabric

MS Shifting Gears from ML to Manual: Moving the Default Semantic Model feature to Graveyard (Part of MS Fabric) Microsoft is making a significant change to its Fabric platform, which will impact how enterprises manage data analytics workflows. The tech giant is retiring auto-generated semantic models, shifting more responsibility from machine learning automation to manual user management. What's Changing? By December 2025, Microsoft Fabric will discontinue its Default Semantic Model feature that automatically creates structured data representations when users set up warehouses, lakehouses, or SQL databases. These semantic models add crucial meaning and context to raw enterprise data, making them essential for analytics workflows. Currently, users can either create custom semantic models or rely on Microsoft's automatically generated defaults. Soon, only the manual option will remain available. Impact on Users This change affects several Fabric features. The Reporting Tab will lose options like 'New Report,' 'Manage default semantic model,' and 'Automatically update semantic model.' Model layouts and context menu shortcuts for quick report creation will also disappear. Existing Default Semantic Models won't vanish but will become standalone entities that enterprises must explicitly manage and maintain themselves, rather than relying on Microsoft's automated processes. Preparing for the Transition Experts recommend starting preparation now. Enterprises should use Fabric Admin APIs to identify existing Default Semantic Models and categorize them as keep, merge, or retire. For models worth preserving, rebuild them as explicit semantic models using Power BI Project (PBIP) or Tabular Model Definition Language (TMDL) to enable proper versioning and governance. Adding metadata through Microsoft Purview can improve discoverability. Training analysts on star schema fundamentals is also crucial. Teams that invest time in cleanup and proper training now will avoid performance issues and messy models later. The Bigger Picture Microsoft frames this change as part of its broader data governance strategy, emphasizing greater user accountability and transparency. While this creates more work upfront, it potentially leads to better-managed, more compliant data assets in the long run.