Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
Azure Databricks released these features and improvements in June 2026.
Note
Releases are staged. Your Azure Databricks account might not be updated until a week or more after the initial release date.
Publishing a Unity Catalog catalog to Microsoft Fabric is in Public Preview
June 18, 2026
You can publish an entire Unity Catalog catalog from Azure Databricks to Microsoft Fabric as a read-only mirrored catalog. Fabric users can query the mirrored Unity Catalog tables directly. See Publish a Unity Catalog catalog to Microsoft Fabric.
Token auto-scoping for personal access tokens is now generally available
June 18, 2026
Token auto-scoping is now generally available. Auto-scoping automatically applies the principle of least privilege to personal access tokens (PATs) by observing which API endpoints a token uses and narrowing its permissions to only those APIs. Auto-scoping is on by default for new long-lived tokens (30 days or longer) and applies to existing tokens with all-APIs access based on historical usage. See Auto-scoping for personal access tokens.
Share Genie Spaces externally using OpenSharing (Beta)
June 17, 2026
Providers can now share a Genie Space with users outside their organization using OpenSharing. Azure Databricks creates a point-in-time snapshot of the space's data assets and instructions. Recipients can mount the share to create a local Genie Space. See Share a Genie Space using OpenSharing.
Redaction of the statement text in the query history system table begins
June 17, 2026
Rolling out starting June 22, the statement_text column in system.query.history returns <Redacted> for users who are not account admins or members of the databricks_pii_access account-level group. Audit any workloads that read statement_text and confirm the principal running them has access to the full SQL text.
See Access redacted statement text.
Omnigent coding agent meta-harness is now in Beta
June 17, 2026
Omnigent is now available in Beta. Omnigent is a coding agent meta-harness that wraps coding agents such as Claude Code and Codex with a common platform: a web UI, persistent and shareable sessions, team collaboration, mobile access, contextual policies, and deployment infrastructure. Swap one line of configuration to switch runtimes while your tools, policies, and prompts stay the same.
OneLake catalog federation is generally available
June 17, 2026
OneLake catalog federation is now generally available. OneLake federation allows you to read data stored in OneLake without copying it, bringing Azure Databricks analytics and AI/BI capabilities directly to your OneLake data. Queries run directly against file storage for better performance and cost-effectiveness. Data access is read-only.
See Enable OneLake catalog federation.
Lakeflow Designer is now generally available
June 16, 2026
Lakeflow Designer is now generally available. Lakeflow Designer is a visual, no-code experience for preparing and transforming data on a drag-and-drop canvas, with all workflows backed by production-ready code governed by Unity Catalog. Lakeflow Designer lets analysts build, preview, and schedule data transformation workflows using built-in operators or natural language.
See Lakeflow Designer.
Lakeflow Designer is coming soon to workspaces with the compliance security profile enabled.
Connect Genie to Microsoft Copilot Cowork (Beta)
June 16, 2026
You can now connect Genie to Microsoft Copilot Cowork through the Genie managed MCP server, so users can ask natural-language questions about their Azure Databricks data without leaving Copilot Cowork. See Genie plugin for Microsoft Copilot Cowork.
Install third-party apps from Databricks Marketplace
June 16, 2026
You can now discover, install, and run third-party data and AI applications from Databricks Marketplace directly in your own Unity Catalog-governed workspace. The provider's code stays closed-source, and you authorize any external access the app needs when you install it. Open Marketplace and filter by the Apps category to get started. See Get access to third-party apps.
Databricks Runtime 19 and Databricks Runtime 19 for Machine Learning are in Beta
June 15, 2026
Databricks Runtime 19 and Databricks Runtime 19 for Machine Learning are now in Beta, powered by Apache Spark 4.2.0.
For release notes, see Databricks Runtime 19 (Beta) and Databricks Runtime 19 for Machine Learning (Beta).
Databricks-to-Databricks Delta Sharing across regulatory domains is now available (Public Preview)
June 15, 2026
Databricks-to-Databricks Delta Sharing supports sharing across regulatory domains. Commercial workspaces can also share with AWS GovCloud, AWS GovCloud DoD, and Azure China. Sharing between Azure China and either AWS GovCloud or AWS GovCloud DoD is not supported. See Databricks-to-Databricks OpenSharing support matrix for cloud environments.
Use Universal Commits to purchase Marketplace partner products
June 15, 2026
You can now use up to 10% of your Universal Commits to purchase eligible third-party partner products through Databricks Marketplace. See Databricks Marketplace Universal Commit Drawdown program.
ai_query is now generally available
June 15, 2026
The ai_query function is now generally available. This general-purpose AI Function lets you query any supported AI model directly from SQL or Python. See Use ai_query.
Share data behind a firewall with SecureConnect (Public Preview)
June 15, 2026
Providers can now use OpenSharing SecureConnect to share data from cloud storage behind a firewall or private endpoint without adding each recipient's network to an allowlist. Recipients access the shares using their existing OpenSharing setup. See Share data behind a firewall with SecureConnect and OpenSharing recipient firewall configuration for SecureConnect.
SecureConnect is also available for clean rooms. See Enable SecureConnect for a clean room.
OIDC is now supported for sharing to Iceberg clients
June 15, 2026
You can now use an Iceberg REST Catalog client, such as Snowflake or OSS Spark, to read OpenSharing shares secured with OpenID Connect (OIDC) federation. For details, see Read data shared using Open ID Connect (OIDC) federation from an Iceberg client and Share with Iceberg recipient.
Configure the recovery period for dropped managed tables (Public Preview)
June 15, 2026
You can now configure the recovery period for dropped Unity Catalog managed tables. By default, tables are recoverable for 7 days after being dropped using the UNDROP TABLE command. You can set the recovery period to 0 hours (to immediately purge dropped data) or between 7–30 days at the catalog or schema level. A longer period provides added protection against accidental drops of critical data, while a shorter period reduces cloud storage costs for workloads that frequently create and drop tables. Requires Databricks Runtime 17.3 and above. See Configure the recovery period.
AUTO CDC flows for streaming tables in Databricks SQL are generally available
June 15, 2026
The FLOW AUTO CDC clause for CREATE STREAMING TABLE in Databricks SQL is now generally available. Use AUTO CDC to process change data capture (CDC) records from a source into a streaming table, automatically handling out-of-order records as SCD type 1 or SCD type 2 without the complex logic that MERGE INTO requires. See CREATE STREAMING TABLE ... FLOW AUTO CDC, generally available.
Email notifications for expiring service principal tokens (Beta)
June 12, 2026
Databricks now automatically sends email notifications to workspace admins when a service principal token is within seven days of expiring. This feature is in Beta. Notifications are sent for tokens with a lifetime greater than seven days that have been used at least once. No configuration is required. See Monitor and revoke personal access tokens.
Managed disaster recovery is now in Public Preview
June 12, 2026
Managed disaster recovery (DR) is now in Public Preview. Managed DR replicates your Unity Catalog metadata, managed table data, and workspace assets to a secondary region, and provides a stable URL that survives failover so downstream clients do not need to be reconfigured. You configure and operate managed DR from the new Resilience section of the account console.
Access to managed disaster recovery is currently gated. Contact your Azure Databricks account team to enable managed DR on your account. See Managed disaster recovery.
Mission Critical add-on is now in Public Preview
June 12, 2026
The Mission Critical workspace add-on is now in Public Preview. The add-on unlocks managed disaster recovery and the existing Enhanced Security and Compliance (ESC) capabilities for a single compute rate, and is enabled per workspace from the new Add-ons tab in the account console.
Access to the Mission Critical add-on is gated. Contact your Azure Databricks account team to enable the add-on on your account.
AI Runtime CLI (Beta)
June 12, 2026
The AI Runtime CLI (air) is now in Beta. Use the air command-line interface to submit and manage distributed training workloads on AI Runtime serverless GPU compute from your local machine, without opening a notebook. See AI Runtime CLI.
Auto time-to-live for automatic row deletion is generally available
June 12, 2026
Auto time-to-live (auto-TTL) is now generally available. Auto-TTL automatically deletes rows from Unity Catalog managed Delta Lake tables, managed Apache Iceberg tables, and streaming tables based on a DATE or TIMESTAMP column. See Automatic row deletion with auto time-to-live.
Converting a partitioned table to liquid clustering is generally available
June 12, 2026
You can now convert an existing partitioned Delta Lake table to liquid clustering with ALTER TABLE ... REPLACE PARTITIONED BY WITH CLUSTER BY. Conversion minimizes reader and writer downtime and supports both external and managed tables. This feature is generally available in Databricks Runtime 18.1 and above. See Convert a partitioned table to liquid clustering.
Improved workspace files read performance
June 16, 2026
For serverless compute and classic compute running Databricks Runtime 18 and above, common file access patterns are now detected for workspace files, and directory metadata and file contents are prefetched in the background. Workloads that read many files, such as find, grep, Python imports, and Git operations, run much faster. See What are workspace files?.
ABAC GRANT policies for models (Beta)
June 11, 2026
ABAC GRANT policies are now in Beta. GRANT policies extend attribute-based access control to dynamically grant Unity Catalog privileges to securable objects whose governed tags match a condition, instead of granting privileges on each object individually. In Beta, a GRANT policy can grant the EXECUTE privilege on models, attached at the catalog or schema level. Both customer-registered MLflow models and Azure Databricks-hosted foundation models in system.ai are supported. See ABAC GRANT policies for models (Beta).
ai_extract is now generally available
June 11, 2026
The ai_extract function is now generally available. This function extracts structured data from text and documents according to a schema you provide, supporting nested objects, arrays, type validation, citations, and confidence scores. See ai_extract function.
For a no-code UI to build, validate, and iterate on extraction agents, see Information Extraction.
ai_classify is now generally available
June 11, 2026
The ai_classify function is now generally available. This function classifies text content according to custom labels you provide, with support for label descriptions, global instructions, and multi-label classification. See ai_classify function.
For a no-code UI to build, validate, and iterate on classification agents, see Classification.
Agent Bricks Supervisor Agent now supports Unity Catalog volumes as subagent tools
June 11, 2026
You can now add Unity Catalog volumes as subagent tools in your Supervisor Agent. See Use Supervisor Agent to create a coordinated multi-agent system.
Schedule deferred policy enforcement for all-purpose compute
June 10, 2026
When you enforce compute policy compliance on out-of-compliance all-purpose compute, you can now schedule the configuration update to apply the next time the compute terminates or restarts, instead of restarting it immediately. Enforce on next restart avoids interrupting running workloads, while Restart and enforce applies the update right away. Workspace admins can cancel a scheduled enforcement for a single compute resource or for an entire policy. See Enforce policy compliance.
Databricks Genie app in Microsoft Teams (Beta)
June 10, 2026
The Databricks Genie app in Microsoft Teams is now available in Beta. The app brings Genie into Microsoft Teams so users can ask data questions and get answers without leaving Microsoft Teams. Chat with the app in a direct message or mention @Databricks Genie in a channel or group chat, and answers route through Genie or a pinned Genie Space. See Databricks Genie app in Microsoft Teams.
Lakehouse Replay (Beta)
June 10, 2026
Lakehouse Replay is now available in Beta. Lakehouse Replay automatically replays a sampled set of read-only serverless workloads from your workspace against upcoming Databricks Runtime releases to detect and fix regressions before they reach production. It requires no setup, configuration, or maintenance. Your data never leaves your workspace, and you are not billed for the compute used to run the replay.
Workspace admins can control the enablement of this feature from the workspace Previews page. See Lakehouse Replay.
Delta Sharing is now OpenSharing
June 10, 2026
Following the release of open-source OpenSharing, Delta Sharing is now OpenSharing. OpenSharing is the new open standard for sharing data and AI assets across platforms and organizations, available both on GitHub and natively on Databricks. See What is OpenSharing? and opensharing.io for more.
Anthropic Claude Fable 5 now available as a Databricks-hosted model
June 9, 2026
Databricks Model Serving now supports Anthropic Claude Fable 5 as a Databricks-hosted model. Claude Fable 5 is built for autonomous knowledge work and coding, designed to handle long-running, complex, and asynchronous tasks with less need for human intervention.
To access this model, use Foundation Model APIs pay-per-token.
Identity federation enabled by default for all new workspaces
June 9, 2026
Over the next couple of weeks, identity federation is being enabled by default for all new Azure Databricks workspaces, including workspaces that are not attached to a Unity Catalog metastore. This change does not affect existing workspaces. See Identity federation.
Databricks Genie app for Slack (Public Preview)
June 9, 2026
The Databricks Genie app for Slack is now available in Public Preview. The app brings Genie into Slack so users can get answers to data questions without leaving Slack. Send a direct message to the app or mention @Genie in a channel or group DM, and answers route through Genie or a specific Genie Space that a channel owner has pinned. See Databricks Genie app for Slack.
The update_flow API for Lakeflow Spark Declarative Pipelines is generally available
June 9, 2026
The @dp.update_flow decorator in Lakeflow Spark Declarative Pipelines lets you add an update flow that writes to a sink in update output mode, emitting only the rows that change in each batch. Unlike append flows, update flows support stateful aggregations without requiring a watermark. See update_flow, generally available.
Reorganized Lakeflow Spark Declarative Pipelines documentation
June 9, 2026
The Lakeflow Spark Declarative Pipelines documentation is reorganized: all conceptual topics are now grouped under Concepts.
Monday.com managed ingestion connector (Beta)
June 8, 2026
The Monday.com connector in Lakeflow Connect is now available in Beta. The connector ingests work management data such as boards, users, teams, and activity logs from Monday.com into Azure Databricks. See Monday.com connector.
Zoho Books managed ingestion connector (Beta)
June 8, 2026
The Zoho Books connector in Lakeflow Connect is now available in Beta. The connector ingests accounting data such as invoices, bills, contacts, and bank transactions from Zoho Books into Azure Databricks. See Zoho Books connector.
External lineage is now generally available
June 8, 2026
External lineage (formerly Bring Your Own Lineage) in Unity Catalog is now generally available. You can register assets that live outside Azure Databricks, such as Salesforce or MySQL sources and Tableau or Power BI dashboards, so that a single lineage graph spans your full data flow. As part of this release, Lakeflow Connect managed ingestion pipelines now automatically record source lineage from source tables to their destination tables in Unity Catalog. See External lineage and Track source data lineage for managed ingestion pipelines.
AI Search supports selected text columns, sorting, and aggregations (Beta)
June 8, 2026
AI Search now supports query-time options for full-text and hybrid queries in Beta. You can limit keyword matching to selected text columns, sort results by columns indexed for sorting, and return aggregate counts for values and ranges. For details, see Search selected text columns, sort results, and return aggregations (Beta).
Foundation model Unity Catalog permissions is generally available
June 5, 2026
Foundation model Unity Catalog permissions is now generally available. This feature requires enablement by your Databricks account team.
Account admins can use Unity Catalog permissions on the system.ai schema to control which Databricks-hosted foundation models your organization can access. Permissions are enforced across pay-per-token, provisioned throughput, and batch inference (AI Functions) workloads.
Managed SharePoint connector adds structured file ingestion and file metadata support (Beta)
June 5, 2026
The managed SharePoint connector in Lakeflow Connect now supports structured file ingestion (CSV, JSON, XML, Excel, Parquet, Avro, ORC), file metadata ingestion, file filters, schema evolution modes, and schema hints. These capabilities replace the previous unstructured-only ingestion approach with a unified managed file source API. See SharePoint connector.
External data access for pipeline streaming tables and materialized views (Public Preview)
June 5, 2026
You can now enable external Delta and Iceberg clients to read streaming tables and materialized views managed by a pipeline through the Unity Catalog and Iceberg catalog REST APIs, without copying the data. See Enable external data access to streaming tables and materialized views, Public Preview.
Improved workspace file write performance on classic compute
June 4, 2026
Workspace file operations on classic compute now use asynchronous flushing by default for Databricks Runtime 17.1 and above. File writes, deletes, and metadata operations are batched and flushed in the background, reducing latency for workloads that write many files, such as Git operations. See What are workspace files?.
Write data from Microsoft Excel back to Azure Databricks
June 4, 2026
Use the Azure Databricks Excel Add-in to write data from Microsoft Excel back to a Unity Catalog table. Create a new table or overwrite an existing table in Azure Databricks without leaving Excel. See Write data back to Azure Databricks using the Azure Databricks Excel Add-in.
Set budgets for Genie usage
June 4, 2026
On July 6, 2026, Genie products will move to a pay-as-you-go pricing model. Each user receives 150 DBUs of free LLM usage every month, equivalent to about $10.50 in the US East region. Usage beyond the free allowance is billed in DBUs. For more information, see what's coming.
To prepare, account admins can now create budgets to monitor and control spending on Genie usage. Scope a budget to Genie to track spend across your account, workspaces, user groups, or individual users, configure email alerts, and set per-user spending limits. See Manage budgets and cost controls for Genie.
Genie Code now supports auto-approve for tool actions
June 4, 2026
Genie Code now offers an auto-approve mode that approves tool actions, like running code or editing notebooks, without prompting you for each request. An AI classifier reviews each action and blocks risky ones. Auto-approve is a productivity feature, not a security boundary. Databricks recommends keeping it off when working with production data or shared resources. See Approve tool actions.
Full page Genie Code (Beta)
June 3, 2026
Full page Genie Code is now available in Beta. The full page experience is a command center for Genie Code, where the active thread is shown prominently, surfacing assets like notebooks and files alongside it as tabs when needed. You can run multiple threads in parallel, switch between them easily, and easily personalize Genie Code with skills, instructions, and MCP servers. See Full page Genie Code.
Table update triggers on Delta Sharing and system tables (Beta)
June 3, 2026
Table update triggers in Lakeflow Jobs can now monitor data shared through Delta Sharing and system tables, in addition to local tables. You can trigger a job whenever a provider updates a shared table or view, or whenever new operational data lands in a system table. This feature supports Databricks-to-Databricks Delta Sharing only. See Add a trigger to OpenSharing and system tables.
Agent Bricks Supervisor Agent now supports web search as a built-in tool
June 3, 2026
You can now enable web search on your Supervisor Agent to answer questions that require current information from the public web. The supervisor returns a synthesized answer with source citations. Web search is also available as a web_search tool in the Supervisor API. See Use Supervisor Agent to create a coordinated multi-agent system.
Pendo connector (Beta)
June 2, 2026
Lakeflow Connect now supports a managed connector for ingestion from Pendo. See Pendo connector.
Zoom Logs managed ingestion connector (Beta)
June 2, 2026
The Zoom Logs connector in Lakeflow Connect is now available in Beta. The connector allows you to ingest user sign-in/sign-out activity logs and admin operation logs from Zoom into Azure Databricks. See Zoom Logs connector.
Managed RabbitMQ connector (Beta)
June 2, 2026
The managed RabbitMQ connector in Lakeflow Connect is now available in Beta. The connector allows you to stream messages from RabbitMQ classic queues into streaming tables in Unity Catalog. See RabbitMQ connector.
Slack Access and Integration Logs managed ingestion connector (Beta)
June 2, 2026
The Slack Access and Integration Logs connector in Lakeflow Connect is now available in Beta. The connector allows you to ingest workspace access logs and app integration change logs from Slack into Azure Databricks. See Slack Access and Integration Logs connector.
Lakebase snapshot storage is now billed
June 1, 2026
Starting June 1, 2026, snapshot storage is billable for Lakebase Autoscaling. Snapshot storage size is shown on the Backup & Restore page. For pricing details, see the Lakebase pricing page. To learn about managing snapshots, see Snapshots.
Vector Search is now AI Search
June 1, 2026
Vector Search has been renamed to AI Search. You can now create full text search indexes without any vectors or embeddings needed. See Databricks AI Search.
Intra-VNet NSG rule hardening for Azure Databricks is now in Public Preview
June 1, 2026
You can now harden the intra-VNet NSG rule to restrict workspace traffic to your host and worker subnet CIDRs instead of the entire virtual network. Applies to VNet injection workspaces on the classic compute plane. See Harden intra-VNet network security group (NSG) rules for Azure Databricks.
Column popularity is now available in the Catalog Explorer
June 1, 2026
Column popularity is now available in the Catalog Explorer UI. To find the most popular columns in a table, go to the table's Overview tab. Popularity is determined by the number of queries against the table that read from a given column. See Popularity.