What's new

New Decision Optimization runtime and CPLEX version (watsonx.ai Runtime)

21 July 2026

Decision Optimization has new options:

• New Decision Optimization runtime. When you run a model in a Decision Optimization experiment, the new do_22.2 runtime is now used by default. See Changing default environments.

• CPLEX V.22.2 is now available in watsonx.ai Runtime. See Model deployment.

Deprecation of notebook sharing by URL in watsonx.ai Studio

21 July 2026

Action required Sharing notebooks by URL is deprecated in watsonx.ai Studio in the Dallas, Frankfurt, London, and Tokyo regions, and will be removed on 8 October 2026. This change strengthens enterprise security, improves account data isolation, and enforces stricter access controls.

Currently, you can generate shareable links to read-only notebook views that anyone with the URL can access, without authentication. Going forward, use more secure sharing options that provide full auditability:

• Share notebooks through project-level collaboration with designated collaborators.
• Export notebooks for controlled distribution.

The removal is phased:

Phase 1: 10 September 2026 — You can no longer create new shareable notebook URLs. Existing shared notebook URLs continue to work during this phase.

Phase 2: 8 October 2026 — All existing shared notebook URLs are permanently deactivated. Anyone who tries to access a previously shared URL receives an error. You retain full access to your notebooks within watsonx.ai Studio, and alternative sharing methods such as project collaboration and controlled access remain available.

View runtime logs with consistent timestamps in DataStage for Cloud Pak for Data

10 July 2026

In DataStage for Cloud Pak for Data, you can now view runtime logs with consistent timestamps and change the displayed time zone in the logs.

Pushdown of processing for definition-based data quality rules

29 June 2026

For definition-based data quality rules that assess data from a Databricks data source, expressions can now be converted to SQL queries and run directly at the data source. For more information, see Pushdown of data quality rules.

Import, enrich, and assess the quality of streaming data

29 June 2026

You can now import metadata from an Apache Kafka data source, enrich that data, and assess its quality. For more information, see Supported connectors for discovery, enrichment, and data quality.

Import lineage from new data sources

26 June 2026

You can now import lineage metadata from the following additional data sources:

• Microsoft Power BI Report Server (Microsoft Power BI Desktop)
• Microsoft SQL Server Integration Services (SSIS)
• Qlik Sense
• Tech preview SAP BusinessObjects

For more information, see Supported connectors for lineage import.

Connect to new data sources by using version 1.2.0-saas of the Manta agent

26 June 2026

You can now import lineage metadata from the following data sources by using the updated Manta agent 1.2.0-saas:

• Apache Hive
• Google BigQuery
• IBM DataStage for Cloud Pak for Data
• IBM Db2
• IBM Db2 for z/OS
• IBM Db2 on Cloud
• Microsoft Power BI (Azure)
• Microsoft Power BI Report Server (Microsoft Power BI Desktop)
• Microsoft SQL Server Integration Services (SSIS)
• Qlik Sense
• SAP BusinessObjects
• Tableau

For more information, see Configuring agents for lineage metadata import.

Deprecation of IBM Data Virtualization as a deep enforcement solution on Cloud Pak for Data as a Service

23 June 2026

IBM Data Virtualization on Cloud Pak for Data as a Service is deprecated and will be removed in October 2026. IBM Data Virtualization is available on-prem through IBM Software Hub and Cloud Pak for Data. See Data Virtualization on Software Hub.

Compare and visualize scenario differences in a Decision Optimization experiment

17 June 2026

You can now compare and visualize differences and similarities between two scenarios in a Decision Optimization experiment. By comparing scenarios, you can better understand how different model formulations, data, and parameters impact your optimization results.

For more information, see Comparing scenarios in a Decision Optimization experiment.

Updates for DataStage for Cloud Pak for Data

05 June 2026

For Presto and watsonx.data Presto, you can now:

• Configure connectors as targets
• Use Insert, Update, and Merge write operations, as well as run Update statements to update data in a table

Experiment with TLA rules in SPSS Modeler by using the Rule lab

29 May 2026

You can now validate and refine your text link analysis (TLA) rules in the Rule lab before applying them to your complete dataset. The Rule lab is an interactive testing environment within the Text Analytics Workbench where you can enter sample text and see how your existing TLA rules match patterns in the sample. When you find patterns that work, you can automatically generate new TLA rules based on the simulation results. With this iterative approach, you can perfect your rules on small samples to save time and improve accuracy before processing large datasets.

For details, see Rule lab.

Scikit-learn library in watsonx.ai Runtime upgraded to version 1.5.2

27 May 2026

The scikit-learn library used in watsonx.ai Runtime was upgraded to use version 1.5.2. The upgrade addresses security vulnerabilities and ties in with other updates. Because of the upgrade, older versions of the scikit-learn library are deprecated and will be removed on 28 June 2026.

Older versions of the scikit-learn library are now deprecated for SPSS Modeler

26 May 2026

The SPSS Modeler runtime was upgraded to use the scikit-learn library version 1.5.2. The upgrade addresses security vulnerabilities and ties in with other updates to SPSS Modeler. SPSS Modeler is one the feature offerings included in watsonx.ai Studio.

Because of the upgrade, older versions of the scikit-learn library are deprecated and will be removed on 28 June 2026.

This change impacts models that were built with older versions of the scikit-learn library, and it specifically affects models built with the HDBSCAN, KDE, Random Forest (RF), Auto Classifier, and Auto Numeric nodes.

• HDBSCAN and KDE models built with the older versions are not supported. These models fail during scoring, and you'll see error messages in SPSS Modeler. You must rebuild these models by using version 1.5.2.
• Random Forest models built with a version older than scikit-learn library version 1.1.1 are not supported. These models fail during scoring, and you'll see error messages in SPSS Modeler. You must rebuild these models by using version 1.5.2.
• Random Forest models built with a scikit-learn library version between 1.1.1 and 1.5.2 can still be scored, but you'll see warnings in SPSS Modeler. You need to be rebuild these models by using version 1.5.2 if you want to continue using the models.
• Auto Classifier and Auto Numeric nodes are also affected if you selected Random Forest as one of the models in the Expert settings.

Additionally, some specific parameters for nodes are different in scikit-learn library version 1.5.2.

• For KDE nodes, matching and kulsinski were deprecated for the metric parameter.
• For Random Forest nodes, the max_features parameter is now sqrt by default.

Optimize IBM Master Data Management matching algorithms with enhanced pair analysis and recommendations

20 May 2026

You can now create and manage multiple pair analysis tasks simultaneously and use their results to generate tuning recommendations for your matching algorithm. With these enhancements, you now have greater flexibility in your matching and algorithm tuning workflow. Additionally, you can now:

• Generate algorithm tuning recommendations before data stewards complete all pair reviews.
• Include manual stewardship decisions (such as link, unlink, and potential match decisions) in the calculations that generate your tuning recommendations.
• Visualize tuning outcomes by using confusion matrices, histograms, and pie charts to understand how changes will impact your algorithm before you apply them.
• Request new pair analysis tasks while others remain in progress.
• Delete unnecessary pair analysis tasks or results.

For details, see Customizing and strengthening your matching algorithm.

Use new attribute to create Data Refinery flows externally without the UI

18 May 2026

You can now use the new shaperAPICreated attribute to create Data Refinery flows programmatically without needing to use the UI. This capability means that you can:

• Use external APIs to create Data Refinery flows.
• Use third-party integrations to generate flows with shaping operations.
• Use automated workflows to create data transformation pipelines.
• Use custom applications to build Data Refinery flows without using the UI.

For more information, see the API documentation: hostname:port_number/v2/data_flow_spark/docs/swagger/index.html

Create Data Refinery flows in folders

18 May 2026

You can now create Data Refinery flows in folders or save existing flows in folders. The information panel shows the folder paths for the flow and the target that you chose. You can also create jobs in folders and modify the flow and target folder paths in the flow settings.

For more information, see Managing Data Refinery flows.

Define parameters for source and target data in Data Refinery flows

18 May 2026

A new parameter step is now available in the job creation wizard for { site.data.keyword.data_refinery }} flows. You can define parameters for both source and target data so that the same job can be used with different data sets. You can also edit existing jobs to use parameters to define source and target data.

For more information, see Creating jobs in Data Refinery.

Cancel Data Refinery jobs in "starting" state

18 May 2026

You can now cancel Data Refinery jobs that are in the Starting state. This enhancement improves job management and resource control.

New connections for Data Refinery

18 May 2026

You can now use the following connections with Data Refinery:

• Vertica
• Microsoft Azure Databricks

For more information, see Supported data sources for Data Refinery.

Enable Skip After SQL on job abort to prevent After SQL from running when a job fails in DataStage

15 May 2026

You can now use the new Skip After SQL on job abort property in the IBM watsonx.data Presto Connector to prevent After SQL statements from running when the main pipeline fails. This option prevents post-processing logic from running if a job does not complete successfully.

Export data lineage to the OpenLineage format

15 May 2026

You can now export data lineage to the OpenLineage format. The exported lineage is saved as .json files that you can use in any third-party platform that supports the OpenLineage standard. For more information, see Exporting data lineage.

Restore your IBM Master Data Management service to its original settings

8 May 2026

You can now reset your IBM Master Data Management service to remove all data, configuration, and history from your instance. Reset your service when you need to start fresh with a clean slate, similar to performing a factory reset on a mobile device. After the reset completes, you can rebuild your master data environment by defining data types, adding data assets, configuring matching algorithms, and completing other key setup tasks.

For more information, see Resetting your IBM Master Data Management service.

Compare two lineage graph versions

01 May 2026

You can now compare two versions of a lineage graph to see how your data flows changed over time. Analyze which assets were added, removed, or modified, and understand the downstream impact of those changes.

For more information, see Comparing lineage versions.

Create assets from data sources that you use for lineage import

01 May 2026

You can now create catalog assets by connecting to the following lineage-specific data sources:

• Informatica PowerCenter
• InfoSphere DataStage
• Microsoft Power BI (Azure)
• MicroStrategy
• Statistical Analysis System (SAS)
• Talend

For more information, see Importing assets with data lineage connections.

Import, enrich, and assess the quality of data from additional data sources

30 April 2026

You can now import metadata from IBM Planning Analytics, enrich that data, and assess its quality.

For details, see Supported connectors for discovery, enrichment, and data quality.

Automate relationship creation in your master data by defining discovery rules

24 April 2026

You can now configure discovery rules to automatically establish and maintain relationships between your master data records and entities in IBM Master Data Management. Configure conditions and filters that evaluate your master data to discover and create relationships. Discovery rules work across record-to-record, record-to-entity, and entity-to-entity relationships, ensuring that your relationships stay current and consistent.

For details, see Discovering relationships in your master data.

Find and fix MDM data quality issues more easily

24 April 2026

You can now search for and resolve data quality issues in your master data by using the Stewardship tab in IBM Master Data Management. Review potential matches that need manual linking decisions and potential overlays that might indicate incorrect record updates. Click an issue to start working on the remediation task. The new streamlined view helps you focus on the issues that need attention so that you can maintain more accurate master data.

For more information, see Completing data stewardship tasks in IBM Master Data Management.

Organize MDM data quality issues with custom master data tags

24 April 2026

You can now create and manage custom tags in IBM Master Data Management to help your team quickly find, identify, and prioritize data quality issues in your master data. Tags work as searchable, color-coded labels that make it easier to organize and track issues across your organization. You can also configure potential overlay workflows to automatically apply tags to the issues they create, streamlining your data stewardship processes.

For more information, see Managing master data tags in IBM Master Data Management.

Modernize your master data management by migrating from InfoSphere MDM

24 April 2026

You can now migrate your existing master data and matching algorithms from IBM InfoSphere Master Data Management (InfoSphere MDM) to the IBM Master Data Management service. By migrating, you gain access to modern, cloud-native capabilities and integration with other Cloud Pak for Data services.

The IBM Master Data Management migration service provides easy-to-use APIs that preserve your data structure and integrity while minimizing downtime. Your master data entities, relationships, groups, and matching algorithms remain intact throughout the migration process. During migration, both systems run in parallel so that you can validate everything before the final cutover.

For more information, see Migrating data into IBM Master Data Management.

Configure user access to master data

13 April 2026

You can now configure how IBM Master Data Management controls user access to data. Access control strategies ensure that only authorized users can access sensitive or confidential information, such as personally identifiable information (PII).

Configure one or both of the following access control types:

• Attribute-based access control (ABAC): Protects specific data characteristics, across all data, from unauthorized users.
• Token-based access control (TBAC): Uses security tokens to define user access at the row level for each record.

For details about setting up data access control, see Configuring access to data in IBM Master Data Management.

Control master data entity attribute composition at the field level

13 April 2026

When configuring attribute composition rules in IBM Master Data Management, you can now define filtering and prioritization logic at the field level by using value-based rules. As a result, you now have finer control over which record attribute values get surfaced to the entity.

Value-based composition rules help you filter low-quality data and construct more accurate, business-aligned entities. You can exclude invalid values like placeholders or dummy data, prioritize specific values in custom order, apply comparator functions to select for conditions such as the longest name or highest score, and create conditional rules that adapt depending on data conditions.

For details, see Defining attribute composition rules in IBM Master Data Management.

Updates for DataStage for Cloud Pak for Data

10 April 2026

You can now select stages and connectors from categories based on their type and function in your DataStage flow.

You can now add your assets directly to the DataStage canvas with the Add asset to canvas + feature.

Monitor processed OpenLineage events

10 April 2026

You can now monitor OpenLineage events in a centralized dashboard to verify event ingestion, identify failed or pending events, and troubleshoot processing issues. The dashboard also helps you understand the overall health of the OpenLineage processing pipeline by showing event volume and trends over time. For more information, see Monitoring OpenLineage events.

Updates for DataStage for Cloud Pak for Data

3 April 2026

Create more connections for StreamSets flows

27 March 2026

You can create more connections to data sources using StreamSets while maintaining common connectivity connections.

You can use the existing FTP (remote file system) connection with StreamSets flows.

Resynchronize data in Relationship explorer

26 March 2026

Administrators can now resynchronize data to ensure that the latest updates are displayed in Relationship explorer. The resynchronization is useful to quickly correct out-of-sync data issues. Additionally, when Relationship explorer is enabled in environments where large volumes of assets and governance artifacts already exist, running resynchronization indexes all data, which then can be discovered.

For details, see Resynchronizing assets and artifacts in the knowledge graph.

Control how long IBM Master Data Management retains historical data

24 March 2026

You can now set the retention period for historical master data to manage memory usage more efficiently. By configuring how long the system keeps historical events, you can control storage costs while maintaining the data history you need for compliance and analysis.

For more information, see Configuring history tracking.

Updates for DataStage for Cloud Pak for Data

20 March 2026

You can now use the AWS Databricks connector.

You can now set a Query data size feature for the following connectors:

• IBM Db2 on Cloud
• IBM Db2
• IBM Db2 for z/OS
• IBM Data Virtualization

View asset ownership and tags directly on the lineage graph

19 March 2026

You can now see tags and the assigned owner, whether a user or a group, displayed directly on assets in the lineage graph. Previously, these details were available only in the metadata details panel. This streamlined view saves time by providing immediate visibility into asset ownership and categorization as you explore data lineage relationships.

Prioritize rare matches over common ones in your master data

18 March 2026

You can now configure your IBM Master Data Management matching algorithm to score matches based on how common or rare the matched values are in your actual dataset. The algorithm uses your real data distribution to boost scores for distinctive matches and reduce scores for common ones.

For example, matching on the rare last name "Xylander" should score higher than matching on the common name "Smith," because rare matches are more likely to identify the same person. This prevents the algorithm from over-scoring matches on common values like "John Smith" while under-scoring matches on distinctive values like "Hamish Xylander."

For more information, see Matching algorithms in IBM Master Data Management.

Updates for DataStage for Cloud Pak for Data

13 March 2026

You can now configure proxy settings for OAuth2 token requests for the Google BigQuery connector.

Revised Dates for IBM Runtime 24.1 Deprecation

12 March 2026

IBM Runtime 24.1 is deprecated. Beginning 11 June 2026, you will no longer be able to create new notebooks, custom environments, or new deployments by using software specifications that are based on the 24.1 runtime. This change applies to watsonx.ai Studio on Cloud Pak for Data as a Service and watsonx as a Service.

To continue creating and deploying assets without disruption and to benefit from the latest capabilities, performance improvements, and security updates, you must transition to IBM Runtime 25.1.

Migrate your assets and workflows to Runtime 25.1 as soon as possible to avoid any impact to your development and deployment processes.

Support for IBM Runtime 24.1 in watsonx.ai Runtime and watsonx.ai Studio will be removed on 9 Jul 2026.

For information on how to transition to Runtime 25.1, see:

• Changing notebook environments
• Managing frameworks and software specifications

Updates for DataStage for Cloud Pak for Data

6 March 2026

You can now use the following connectors:

• AlloyDB for PostgreSQL connector
• Microsoft SharePoint Files connector

The Data Replication beta is deprecated

6 March 2026

The beta version of the Data Replication service is now deprecated and will be removed on 13 March 2026.

New ONNX software specification for deploying ML models

6 March 2026

The onnxruntime_opset_21 software specification for deploying machine learning models is now available. It provides enhanced performance and compatibility with the latest ONNX model formats.

For more information, see Supported software specifications.

Export data lineage to Collibra

6 March 2026

You can now export data lineage and view it in Collibra. If you transfer lineage information into Collibra data governance platform, you can see a comprehensive view of your data flows and dependencies within your governance framework. For more information, see Exporting data lineage to Collibra.

Filter relationships on the relationship explorer by their state

6 March 2026

When you view data in the relationship explorer, you can now filter relationships by state: published or draft. Published relationships are final, governed relationships that are available across the organization. By filtering out draft relationships, you can focus on the approved version of each relationship.

Use Apache Spark 3.5 and Apache Spark 4.0 to run notebooks and scripts

26 February 2026

These runtimes are now supported for notebooks and R scripts in projects:

• Default Spark 3.5 & Python 3.11
• Default Spark 3.5 & R 4.3
• Default Spark 4.0 & Python 3.11
• Default Spark 4.0 & R 4.3

For details on available notebook environments, see Compute options for the notebook editor.

Stream change events from your master data to downstream systems

18 February 2026

Now IBM Master Data Management can, in real time, propagate changes in your record and entity data directly to downstream systems through a connected Apache Kafka server. Streaming ensures that your users and systems always have the freshest and most up-to-date master data. You can configure master data streaming only through the IBM Master Data Management API.

For more information, see Streaming record and entity data changes.

Updates for DataStage for Cloud Pak for Data

13 February 2026

You can now change the queue priority at a flow, job, and project level on remote engine projects.

Compare master data records side by side

13 February 2026

You can now compare two records in a split view to determine whether they belong together in an IBM Master Data Management entity. Data stewards can explore and compare records and entities, view similarity analysis, and make informed decisions about entity composition.

For details, see Analyzing and comparing master data.

Updates for DataStage for Cloud Pak for Data

6 February 2026

New data sources for lineage metadata import

5 February 2026

You can now import lineage metadata from the Informatica PowerCenter and Talend data sources. After the data is imported, you can visualize it on a lineage graph. For more information, see Supported connectors for lineage import.

Import, enrich, and assess the quality of data from additional data sources

06 February 2026

You can now import metadata from Amazon Aurora for PostgreSQL, enrich that data, and assess its quality.

In addition, you can now write analysis output to tables in Db2 for i.

For details, see Supported connectors for discovery, enrichment, and data quality.

Data Refinery is available in the AWS Mumbai region

30 January 2026

Data refinery is now available in the Mumbai (AWS) region. You can select Mumbai as your preferred region when you log in. You can run Data Refinery jobs with the new Default Data Refinery with Spark 3.4 & R 4.3 environment.

Automatically add Personal Information and Sensitive Personal Information classifications

30 January 2026

When you generate business terms, the Personal Information and Sensitive Personal Information classifications are automatically added where applicable. You can use these classifications to control groupings of assets in your company and protect highly sensitive data. For more information, see Generating business terms, Classifications, and Predefined classifications.

Import lineage metadata from any deployment type of IBM Cognos Analytics and IBM DataStage for Cloud Pak for Data

30 January 2026

When you import lineage metadata from IBM Cognos Analytics and IBM DataStage for Cloud Pak for Data, you can now connect to any deployment type of these technologies that you can access over the network. When you configure the connection, you now specify a deployment type with the new required property. For more information, see Creating a connection to IBM Cognos Analytics and Creating a connection to IBM DataStage for Cloud Pak for Data.

New data sources for lineage metadata import

29 January 2026

You can now import lineage metadata from the Statistical Analysis System (SAS) and Apache Hive data sources. After the data is imported, you can visualize it on a lineage graph. For more information, see Supported connectors for lineage import.

IBM Runtime 24.1 Deprecation

27 January 2026

IBM Runtime 24.1 is deprecated. Beginning 12 March 2026, you cannot create new notebooks or custom environments by using 24.1 runtimes. Also, you cannot create new deployments with software specifications that are based on the 24.1 runtime. To ensure a seamless experience and to leverage the latest features and improvements, switch to IBM Runtime 25.1. Support for IBM Runtime 24.1 in watsonx.ai Runtime and watsonx.ai Studio will be removed on 16 April 2026. For more information, see:

• Changing notebook environments
• Managing frameworks and software specifications
• Migrating custom environment templates from Runtime 24.1 to Runtime 25.1

Publish data quality rules to catalogs

23 January 2026

You can now publish data quality rules to catalogs.

For details, see Managing data quality rules.

Planned deprecation of App ID

21 January 2026

As of 25 February 2026, the App ID federated login option will be replaced with IBM Cloud SAML service provider. If you're currently using AppID, you can continue using it, but it will not be possible to add new App ID instances after 25 February 2026. We encourage you to switch to IBM Cloud SAML.

For more information, see Federating with the IBM Cloud SAML service provider (SP).

New data asset selection window in Decision Optimization experiments

19 January 2026

You can now import data to your Decision Optimization scenario and to your project from the Prepare Data view in one step. When you click the new button Add data, a new window opens where you can select files from your project or local files. Local files that you upload will be also added to your project. You no longer have to first import files to the project and then import them into your scenario. For more information, see Adding data to a scenario.

Updates for DataStage for Cloud Pak for Data

16 January 2026

You can now use the AudienceURL feature for the Salesforce API for DataStage connector.

You can now use the LDAP external authentication type for the MongoDB connector.

You can now set log retention value for migrated optimized pipeline jobs.

Updates for DataStage for Cloud Pak for Data

09 January 2026

You can now use the Microsoft Sharepoint Lists connector as source and target modes.

For the DataStage Assistant, you can now predict non-linear flows that contain branching paths.

For the DataStage Assistant, you can now summarize your flows.

For the DataStage Assistant, you can now generate a name for a stage based on its properties.

Runtime 25.1 is available

16 December 2025

In Watsonx.ai Studio, you can now write your Python and R code in the new 25.1 runtime

In the notebook editor, you can use:

• Runtime 25.1 on Python 3.12
• Runtime 25.1 on R 4.4
• NLP + DO Runtime 25.1 on Python 3.12
• GPU V100 Runtime 25.1 on Python 3.12
• GPU 2xV100 Runtime 25.1 on Python 3.12

For more information, see Compute options for the notebook editor.

In Watsonx.ai Runtime, you can now use the software specifications that are based on the latest 25.1 runtime to:

• Deploy Python code that is based on Python 3.12
• Deploy models that are based on the latest versions of the most important machine learning frameworks
• Run code that uses packages that are dedicated for generative AI

For more information, see Frameworks and software specifications.

IBM Match 360 as a Service is now known as IBM Master Data Management

15 Dec 2025

The IBM Match 360 service is renamed to IBM Master Data Management.

View historical data for entities, records, and relationships in IBM Master Data Management

15 Dec 2025

You can now view the history of each entity, record, and relationship in your master data and compare historical attribute values to the current version. Select any past update to view the attribute values at that point in time and also see whether each update was initiated by a user, source system, or linkage action. You can use this capability to help with audit tracking and analysis of data changes over time.

A data engineer can configure whether the service keeps historical data. Storing history details increases the storage requirements of your database.

For details, see Configuring history tracking and Exploring historical master data.

Updates for DataStage for Cloud Pak for Data

12 December 2025

You can now use CPDCTL with enabled proxy flag on the remote engine.

For the Snowflake connector, you can now specify other file format options for the Parquet file during the bulk load mode.

Merge child assets into parent on the lineage graph to simplify the view

11 December 2025

When you view a complex lineage graph, you can now easily hide all child assets of a particular parent asset by using the Merge into parent option. Before the assets are hidden from the view, when you hover over the option name, all child assets are highlighted and their number is displayed. Merge assets into parent to display a more general view of your data. For details, see Managing data lineage graph.

Explain data quality rules with AI

11 December 2025

You can now use AI to automatically generate plain English descriptions for data quality rules and keep those descriptions up-to-date when the logic of a rule, such as expressions, bindings, or SQL statements, changes. By default, this option in enabled for all existing and new products, but you can disable this capability for any project where you don't need it. If you want to have AI-generated descriptions for data quality rules that you created prior to 11 December 2025, you must manually update those rules.

For more information, see Data quality settings.

Updates for DataStage for Cloud Pak for Data

05 December 2025

For the Snowflake connector, you can now reject records with the bulk load and bulk load merge options.

Include metadata details when exporting lineage graphs to PDF

04 December 2025

When you export the lineage graph to a PDF file, you can include metadata details in the exported file, such as description, tags, and others. For more information, see Downloading a graph.

New and enhanced features in Data Virtualization

03 December 2025

View more information about group types and hierarchy types in IBM Master Data Management

24 Nov 2025

When you open a hierarchy type or group type from the IBM Master Data Management data types page, you can now see more details at a glance, including who created the type, when they created it, and a list of group or hierarchy instances that are based on the selected type. You can also navigate directly from the data types page to the Workspace view to manage each hierarchy and group instance and its members.

For details, see Customizing your data types.

Python 3.12 is now available in Decision Optimization experiments

21 November 2025

In addition to Python 3.11, you can now use Python 3.12 in your Decision Optimization environment to run and deploy Decision Optimization models that are formulated in DOcplex in Decision Optimization experiments. Modeling Assistant models also use Python because DOcplex code is generated when models are run or deployed.

To update your environment, see Configuring environments.

To update existing deployed models, see Changing Python version for an existing deployed model with the REST API.

Create more connections for StreamSets flows

20 November 2025

You can create more connections to data sources using StreamSets while maintaining common connectivity connections.

The following are existing connections that can use StreamSets flows:

• Google BigQuery
• Google Cloud Pub/Sub
• MongoDB

The following are new connections that only work with StreamSets flows:

• JMS

Updates for DataStage for Cloud Pak for Data

21 November 2025

You can now use Sparse lookup property with the Snowflake connector.

Updates for DataStage for Cloud Pak for Data

14 November 2025

When you add the Snowflake connector as a target, you can now specify a Maximum file size property to define the size of the staging area file in MB.

You can now use the OAuth authentication method for the Snowflake connector.

You can now use the Login timeout property for the Datastax connector.

You can now use the Bulk load merge write mode for the Snowflake connector.

You can now use the Load write mode for the Snowflake connector.

You can now specify alternate servers for the Cassandra connection.

You can now use parquet and delimited file format for the Snowflake connector.

You can now specify storage integration name for the Snowflake connector.

Replicate range-partitioned Db2 tables to supported target data stores

12 November 2025

You can now use the Data replication service to replicate Db2 tables that are partitioned based on the range of values in one or more columns. These types of tables are also known as range-partitioned tables. You can replicate range-partitioned tables to Db2 on Cloud, Db2 Warehouse, watsonx.data, and Apache Kafka target data stores.

For details, see Replicating IBM Db2 data and Replicating IBM Db2 on Cloud data.

Starting parents are introduced in the data lineage graph

07 November 2025

When you select an asset to be a starting asset in the lineage, all assets that are higher in the hierarchy are marked as starting parents. Also, all child assets of the selected asset are marked as starting assets. This distinction clarifies which assets are selected as the starting points for the lineage. For more information, see Viewing data lineage.

Create SQL-based data assets and data quality rules from text queries

07 November 2025

When you create an SQL-based data asset or an SQL-based data quality rule, you can now enter your query in plain text and have this text query converted to SQL for actually creating the asset. The project in which you are working must be enabled for natural language queries.

For more information, see Creating data assets by using SQL queries, Creating SQL-based data quality rules, and Data intelligence tools settings.

Tech preview This is a technology preview and is not yet supported for use in production environments.

Override storage type in SPSS Modeler

07 November 2025

You can now override the inferences that the Data Asset node makes about storage types when it imports data. Previously, the Data Asset node would read a sample of the data that it imports to infer what type of data is in a field of a table, such as an integer, date, or string. Changing the storage type after this inference was difficult. Now you can easily override the storage type for a field in the settings for the Data Asset node.

For more information, see Data Asset node.

Configure potential match workflows for each entity type (IBM Master Data Management)

30 October 2025

You can now configure a potential match workflow for each entity type in your data model from the Task types page in IBM Master Data Management. Potential match workflows identify matching issues within your data, then create and assign tasks for data stewards to resolve them.

For details, see Configuring a potential match workflow.

Force stop replication jobs from the UI

24 October 2025

You can now forcefully stop replication jobs that are stuck in the "Stopping" state. Jobs stuck in this transitional state might block new replication jobs. The force stop capability cleans up replication-related system resources.

For details, see Managing replication jobs.

Updates for DataStage for Cloud Pak for Data

24 October 2025

You can now run test connection in the remote engine environment.

You can now import or export data definitions in the new Transformer design.

You can now open a Lookup stage faster with a large number of input links in a flow design.

Edit any attribute of an IBM Master Data Management entity, including composite values

24 October 2025

You can now edit any of a master data entity's attribute values, even if the original value was derived from the entity's member records by applying attribute composition rules. When a data steward overrides a composite value, the user-defined value will be maintained even if the composition of the entity changes.

This capability is only available for entity types that a data engineer has configured to enable entity persistence.

For details, see Adding and editing records and entities.

Navigate and filter lineage data more easily

16 October 2025

Viewing data on lineage is enhanced in the following ways:

• You can filter assets by the quality score.
• You can expand or collapse all asset children in the initial lineage graph view with just one click.
• When you split an asset to display its child assets, these child assets are highlighted for better visibility.
• When you select a node, all its related paths are highlighted to better see where data flows.
• When you select a node, you can easily hide assets that are related in the upstream or downstream direction, or assets that are not related to the selected node.

For information about lineage, see Data lineage.

Create more connections for StreamSets flows

16 October 2025

You can create more connections to data sources using StreamSets while maintaining common connectivity connections.

The following are existing connections that can use StreamSets flows:

• Amazon S3
• Microsoft Azure Data Lake Storage
• Microsoft SQL Server

Updates for DataStage for Cloud Pak for Data

10 October 2025

PostgreSQL driver is now updated to version 6.0.0.1843.

DBT is now updated to version 1.9.7.

You can now use Microsoft Fabric Warehouse connector in DataStage for Cloud Pak for Data.

In Redshift connector, you can now use DELETE write method.

Amazon S3 connector now supports GZIP and uncompressed options for CSV and delimited file format.

In Snowflake and FTP connectors, you can now use encrypted Private Key.

In Teradata database you can now use SSL Protocol property.

File names with spaces from Microsoft OneDrive are now supported.

Db2 jar version is now updated to 4.33.45.

Db2 license jar version is now updated to 4.33.31.

Create mapping rules for advanced interpretation of OpenLineage events for lineage generation

10 October 2025

You can now create mapping rules to define how your custom technologies are interpreted and then represented in lineage. For more information, see Mapping OpenLineage events.

Compare models in Decision Optimization experiments

08 October 2025

You can now compare models from different scenarios in a Decision Optimization experiment and compare the log files when the models are solved. Comparing models in this way is useful as you can see the different scenarios side by side. For more information, see Comparing scenario models and log files in a Decision Optimization experiment.

Generate business terms with generative AI-powered glossary

06 October 2025

Automatically generate business terms to accelerate the development of glossaries and enhance data governance efficiency. New business terms are generated from database sources based on a specific business need. The terms have defined names and descriptions. Also, they might automatically have assigned relationships to other business terms in the glossary. For more information, see Generating business terms.

Connect to new data sources by using the Manta agent

3 October 2025

You can now import lineage metadata from the following data sources by using an agent:

• Amazon RDS for PostgreSQL
• Amazon Redshift
• Greenplum
• IBM Cloud Databases for PostgreSQL
• PostgreSQL

You can install a Manta agent on a system with direct access to a data source if a direct connection from your IBM instance is not possible. For more information, see Configuring agents for lineage metadata import.

Customize lineage visualization with enhanced filters

3 October 2025

You can now adjust the initial view of your lineage by using advanced filters. Decide on the scope of data and type of assets that you want to see on your lineage.

For more information, see Viewing data lineage.

Analyze Chinese text data in SPSS Modeler with Text Analytics

2 October 2025

You can now use the Text Analytics nodes in SPSS Modeler, such as the Text Link Analysis node and Text Mining node, to analyze text data written in Simplified and Traditional Chinese. Text Analytics nodes use advanced linguistic technologies and text mining techniques to analyze text data and extract concepts, patterns, and categories.

For more information about Text Analytics, see Text Analytics.

Create connections for StreamSets flows

2 October 2025

You can create more connections to data sources using StreamSets while maintaining common connectivity connections.

The following are existing connections that can use StreamSets flows:

• Google Cloud Storage
• MySQL

Updates for DataStage for Cloud Pak for Data

19 September 2025

DB2JCC4 driver is now upgraded to 4.33 version. With the new version, you can use security settings such as encrypted password for the DB2 connectors.

Backward compatibility with cpdctl binary version 1.7.10 is now fixed. You can now run pipeline jobs in optimize mode without failing.

Sparse lookup is now available for IBM-Cloud and AmazonRDS PostgreSQL.

Resolved the issue where a parameter of string type was handled as encrypted type which was causing run failure.

Folders for project assets are generally available

18 September 2025

You can now organize your project assets with folders. To use folders in a project, administrators must enable the feature in a project, and then administrators and editors create and manage folders.

See Organizing assets with folders.

Run composite API transactions in IBM Master Data Management

18 Sept 2025

You can now run composite transactions on master data by using the IBM Master Data Management composite_service API endpoint. Composite transactions bundle multiple dependent API calls into a single request to reduce network overhead, centralize error handling, and produce a consolidated response for the entire sequence of operations.

For details, see Running composite transactions in the IBM Master Data Management API Reference documentation.

New Teradata data source for lineage metadata import

12 September 2025

You can now import lineage metadata from the Teradata data source. After the data is imported, you can visualize it on a lineage graph.

For more information, see Supported connectors for lineage import.

Track and record modifications to source databases by using a change log when you replicate data

11 September 2025

You can now configure your Data Replication asset to use the Change Log business goal. With this business goal, you can replicate changes in the source database to the target database and simultaneously maintain a record of all modifications made to entries in the source database.

For details, see Setting a business goal for a Data Replication asset.

Use PostgreSQL as a source data store for replicating data in your project

11 September 2025

You can now replicate data from PostgreSQL with the Data Replication service. For details, see Supported Data Replication connections.

Removal of Federated Learning

11 September 2025

The Federated Learning feature is no longer available.

Enhancements to the IBM Master Data Management service

9 September 2025

Several key capabilities are now enhanced within the IBM Master Data Management service.

Spark 3.4 & R 4.2 runtime deprecation

4 September 2025

As of 4th September 2025, the Default Spark 3.4 & R 4.2 runtime in Watsonx.ai Studio is deprecated. From 3rd October 2025, you cannot create new notebooks or custom environments using Default Spark 3.4 & R 4.2 runtime. You can still run your code that uses the deprecated runtime until 3rd November 2025. To avoid disruption, make sure to update your code to use the Default Spark 3.4 & R 4.3 runtime.

• For information about changing environments, see Changing notebook environments.
• For a list of available Spark environments, see Spark environment templates.

Automate data quality analysis in IBM watsonx.data intelligence

29 August 2025

Have data quality checks automatically generated for your data instead of running a basic data quality analysis with a fixed set of predefined checks in metadata enrichment. Immediately run these suggested checks as another step in the enrichment, or review and adjust the checks, then add the run step to the enrichment before you re-enrich your data.

Data quality checks can be generated based on profiling results, generated based on constraints that are defined in assigned business terms, or you can manually add them to check an entire data asset or specific columns. Users can decide whether these checks should be applied without further review, or review and then have the reviewed ones applied. Available types of data quality checks include the checks that were available in earlier releases and a set of new checks, for example, checks for historical stability and referential integrity.

For details, see:

• Designing metadata enrichment
• Available types of data quality checks

Use definition-based data quality rules with external bindings multiple times in your DataStage flows

29 August 2025

You can now create data quality rules that you can use any number of times in one or more DataStage flows. Create a definition-based rule with external bindings in the project. Then, add the rule to a DataStage flow by selecting it from the Asset Browser stage as many times as needed. When you update the respective data quality rule asset, the changes are automatically reflected in any flow that contains the rule. For each rule that you run, you can decide which associated DataStage flows are run.

For details, see Creating rules from data quality definitions.

Tech preview This is a technology preview.

Export lineage to PDF and CSV with new output options

28 August 2025

You can now export lineage to PDF and CSV files with new output options. With the Assets connections option, the exported file contains all starting assets and connections to unique assets. With the Lineage paths option, the exported file contains all unique paths that come from the starting assets.

For more information, see Downloading a data lineage graph.

Sample data quality definitions are now available in the IBM Cloud Resource hub

28 August 2025

190 data quality definitions are now available as a sample project in the Resource Hub. The sample definitions cover a wide range of standard checks, for example, for validity or completeness of certain values. Use them as-is to build your data quality rules, use them as templates, or learn from them as models.

For more information, see Sample data quality definitions.

Create connections for StreamSets flows

28 August 2025

You can create more connections to data sources using StreamSets while maintaining common connectivity connections.

The following are existing connections that can use StreamSets flows:

• IBM watsonx.data Presto
• Microsoft Azure Blob Storage
• PostgreSQL

The following are new connections that only work with StreamSets flows:

• Databricks Delta Lake
• Splunk

Updates for environments for running Data Refinery flow jobs

22 August 2025

The new Default Spark 3.4 & R 4.3 environment is added. You can now select Default Spark 3.4 & R 4.3 when you select an environment for a Data Refinery flow job.

Import sample predefined business terms with IBM watsonx.data intelligence

7 August 2025

You can now install a sample of 100 predefined business terms by using the Knowledge Accelerators importing UI. Use the sample business terms to classify personal data across key concepts such as Person, Organization, Employment Record, Contact Information, and Finance Information.

For more information, see Getting started with Knowledge Accelerators

Configure potential overlay task workflows in IBM Master Data Management

7 August 2025

You can now configure workflows for potential overlay tasks from the new Task types page in IBM Master Data Management. You can configure the conditions and scenarios that generate potential overlay tasks by choosing which data sources and attributes are monitored, and then defining the degree of change that results in the creation of a potential overlay task. Define as many workflow type configurations as you need to cover all of your organization's potential overlay handling scenarios.

For more information, see Configuring a potential overlay workflow type.

Create connections for StreamSets flows

31 July 2025

You can now create connections to data sources using StreamSets while maintaining common connectivity connections. The following are existing connections that can use StreamSets flows:

• Apache Kafka
• IBM Db2
• Oracle
• Snowflake

The following are new connections that only work with StreamSets flows:

• JDBC
• Web Client

Deprecation of Federated Learning

25 July 2025

The Federated Learning feature is deprecated and will be removed from watsonx.ai and Cloud Pak for Data as a Service on 10 Sep 2025.

New default method for name generation in the metadata enrichment project settings

25 July 2025

When you create a project, the default method for generating display names that is defined in the project settings for metadata enrichment is now Generative AI. For details, see Display name in Default enrichment settings.

Train AutoAI experiments with large data sets using incremental learning

24 July 2025

You can now prepare an AutoAI experiment to train incrementally with batches of data. Each batch is scored separately, so you can review performance of each batch when viewing the results. After training completes, you can continue training the experiment in an auto-generated notebook by supplying more batches of data.

For more information, see Using incremental learning to train with a large data set.

Manta agents can be used on FIPS-enabled environments

18 July 2025

When you import lineage metadata by using external Manta agents, you can now also use these agents on environments that are FIPS-enabled. For more information, see Configuring agent to run on the FIPS-enabled enviroment.

Account resource scoping is now enabled permanently

17 July 2025

Account resource scoping is now permanently enabled. You can access only those projects that were created in the currently selected IBM Cloud account. You must switch to a different account to see projects created in other accounts.

The following options are removed:

• The Resource scope option in account settings. Resource scoping is always enabled.
• The Restrict who can be a collaborator option during project creation. All projects require account membership.

Publish SQL query assets to catalogs

17 July 2025

You can now publish SQL query asset types to catalogs. In governed catalogs, previews of SQL query data assets with data protection rules enforced are also available.

For details, see Publishing assets from a project into a catalog.

Document the lifecycle of a project by using the Document editor

11 July 2025

You can now use the new Document editor to create, edit, and store project documentation, including a project readme file. This dedicated tool helps you keep track of important details, improve collaboration, and maintain organized records of the project in a simple way. You can find the Document editor on the Overview page of a project. To enable the Document editor for older projects, click Migrate to migrate your current readme file to the editor. For more information, see Project documentation and notifications.

Easily upload files for lineage metadata import

27 June 2025

You can now upload files for lineage metadata import directly when you create a new metadata import. For more information, see Designing metadata imports: External inputs.

Extract metadata from on-premises data sources and analyze data lineage with Manta agents

27 June 2025

You can now extract metadata from on-premises data sources and send them for data lineage analysis by using Manta agents. For more information, see Configuring agents for lineage metadata import.

New data sources for lineage metadata import

27 June 2025

You can now import lineage metadata from the following additional data sources. After the data is imported, you can visualize it on a lineage graph.

• IBM Cognos Analytics
• Microsoft Azure Databricks.

For more information, see Supported connectors for lineage import.

Removal of Python 10 for Decision Optimization

26 June 2025

Python 10 is now removed from Decision Optimization. You can use Python 3.11 with the Decision Optimization environment.

Python is used to run and deploy Decision Optimization models that are formulated in DOcplex in Decision Optimization experiments. Modeling Assistant models also use Python because DOcplex code is generated when models are run or deployed.

To update your environment, see Configuring environments.

To update existing deployed models, see Changing Python version for an existing deployed model with the REST API.

IBM Match 360 as a Service is now generally available

24 June 2025

IBM Match 360 as a Service is now generally available in the Dallas and Toronto data centers. IBM Match 360 as a Service is a cloud-native, multi-domain master data management (MDM) solution that empowers organizations to unify and enrich data from across the enterprise, delivering a trusted 360-degree view of your most critical business entities.

The core of IBM Match 360 is its world class matching engine, which data engineers can tune and customize to meet the needs of the organization. With intelligent, machine learning–guided matching and self-service tools, IBM Match 360 helps data engineers and data stewards to streamline data integration, measureably improve data quality, and unlock deeper business insights.

Whether you're modernizing legacy MDM systems or building a data fabric for AI, IBM Match 360 as a Service provides the foundation for trusted, actionable data across your business. Start using IBM Match 360 as a Service today with a free, 30 day Trial plan or with a paid Essential plan.

For more information about IBM Match 360 as a Service, see IBM Match 360 as a Service.

Deploy models that were trained with Spark 3.5

9 June 2025

In watsonx.ai Runtime You can now deploy machine learning models that were trained with Spark 3.5. For details about available software specifications, see Frameworks and software specifications.

Security enhancement for evaluation jobs in data quality workflows

30 May 2025

Evaluations now support Task credentials to enhance security compliance. When processing evaluations from the user interface, additional confirmation is required to help ensure secure access. For new AI assets, the credentials that are provided during onboarding are used to securely connect to external services during evaluation. For existing assets, if no credentials are detected, a banner is automatically displayed on the Evaluations page with guidance on how to resolve the issue.

Multilingual Metrics Support in Generative AI Quality Monitor

30 May 2025

You can now evaluate Generative AI outputs in multiple languages by using the Generative AI Quality monitor in watsonx.governance. With this new feature, you can create both design-time and runtime evaluations using certain selected metrics for summarization, generation, extraction, question answering (QA), and retrieval-augmented generation (RAG) tasks. The languages include: English (en), Japanese (ja), German (de), French (fr), Spanish (es), Arabic (ar), Italian (it), Portuguese (pt), Korean (ko), and Danish (DA). For more information, see Evaluating generative AI output in multiple languages.

Extended Character Limit Support in Prompt Template Assets

30 May 2025

You can now work with significantly larger text data in Prompt Template Assets (PTAs) by using a new column configuration option in the UI to increase the character limit value to more than 32k (excluding content in a Classification Type Task or meta fields). This enhancement is ideal for use cases like long-form summarization, transcript processing, and other large text scenarios in Generative AI workflow.

Create Detached Prompts in Watson OpenScale UI

16 May 2025

You can now create Detached Prompts directly through the user interface in both projects and deployment spaces. Previously, this action was only available through notebooks. This enhancement simplifies Detached prompt creation and improves accessibility for users.

Removing duplicate fields from json responses to deployment-related API calls

5 May 2025

The API calls for creating and managing deployments in watsonx.ai Runtime currently return duplicate fields in entity and metadata object fields. This behavior is now deprecated. Starting with 5th of August 2025, these fields will not be included in the entity object field:

• name
• space_id
• project_id
• tags
• description

To avoid disruption, you must update your code that currently handles deployment-related API responses. For more information, see Deployment samples.

IBM Knowledge Catalog name change

02 May 2025

The IBM Cloud service IBM Knowledge Catalog is renamed to IBM watsonx.data intelligence.

Existing clients of IBM Knowledge Catalog, Manta Data Lineage, and Data Product Hub can continue using their previous plans.

The changes are effective 02 May, 2025.

Users with the Viewer role don't see empty custom attributes

30 April 2025

Users with the Viewer role see only custom attributes that have assigned values.

Finer-grained enrichment settings for the generation of names and descriptions (IBM watsonx.data intelligence)

25 April 2025

In the metadata enrichment settings for the Expand metadata objective, you can now choose to generate display names for assets and columns by using either fuzzy matching or generative AI. You might want to use fuzzy matching, for example, to work with a domain-specific business vocabulary.

You also now have the option to turn off description generation.

For details, see Metadata enrichment default settings.

Run the full set of curation and data quality capabilities on more data sources and data formats (IBM watsonx.data intelligence)

25 April 2025

You now have additional options when curating data from the following data sources: Amazon RDS for MySQL : You can now run data quality rules on assets from Amazon RDS MySQL databases, create query-based data assets, and also write analysis output to such databases.

IBM Cloud Object Storage

You can now import, enrich, and run definition-based data quality rules on tables in Delta Lake and Iceberg format from IBM Cloud Object Storage.

For details, see Supported data sources for curation and data quality

Manta Data Lineage is now also available in the Tokyo region

11 April 2025

Manta Data Lineage is now also available in the Tokyo data center. You can select Tokyo as your preferred region when you sign up.

For more information about product features that are available in the Tokyo region, see Regional availability for services and features.

Manta Data Lineage is now also available in the London region

4 April 2025

Manta Data Lineage is now also available in the London data center. You can select London as your preferred region when you sign up.

For more information about product features that are available in the London region, see Regional availability for services and features.

Deprecation of data location and sovereignty rules for Data Privacy

25 March 2025

The data location and sovereignty rules were experimental features that provide attributes-based access control of data assets based on their location or sovereignty. These data location and sovereignty rules are now deprecated.

If you have questions or concerns that are related to the deprecation, you can open a support ticket.

Manta Data Lineage is now also available in the Frankfurt region

21 March 2025

Manta Data Lineage is now also available in the Frankfurt data center. You can select Frankfurt as your preferred region when you sign up.

For more information about product features that are available in the Frankfurt region, see Regional availability for services and features.

Write analysis output to more databases (IBM watsonx.data intelligence)

20 March 2025

You can now write the analysis output of advanced profiling or running data quality rules also to Amazon RDS for Oracle or Amazon RDS for PostgreSQL databases.

For more information, see Supported data sources for curation and data quality.

Enhancements to output tables for data quality rules (IBM watsonx.data intelligence)

20 March 2025

You can now specify certain parameters to generate dynamic names for your rule output tables. Also, you can now select whether you want such tables to be added to your project. They are no longer automatically added.

For consistent setup of rule output tables, a project administrator can now configure default settings for a project. You can still overwrite these settings for individual rules.

For details, see Configuring output settings for data quality rules and Project settings for data quality.

New and enhanced features in Data Virtualization

17 March 2025

The following new and enhanced features are available in Data Virtualization:

Removed the Recommended assets tab (IBM watsonx.data intelligence)

14 March 2025

The Recommended tab with assets that were recommended to you based on the properties common to the assets that you viewed, created, and added to projects is no longer available.

Gen AI based metadata enrichment features are now available in all regions (IBM watsonx.data intelligence)

14 March 2025

The options for semantic and AI-augmented data enrichment in IBM watsonx.data intelligence are now also available in the Frankfurt, London, and Tokyo regions.

Run additional enrichment and data quality capabilities on more data sources (IBM watsonx.data intelligence)

14 March 2025

You now have additional options when enriching or running data quality rules on the following data sources:

Amazon RDS for Oracle

Enrich metadata and create query-based data assets from Amazon RDS for Oracle databases.

Amazon RDS for PostgreSQL

Run data quality rules on assets and create query-based data assets from Amazon RDS for PostgreSQL databases.

Snowflake

Write analysis output to a Snowflake data storage.

For more information, see Supported data sources for curation and data quality.

New Google BigQuery data source for lineage metadata import

6 March 2025

You can now import lineage metadata from the Google BigQuery data source. After the data is imported, you can visualize it on a lineage graph.

For more information, see:

• Supported data sources for curation and data quality
• Google BigQuery

Code snippets are now available in Decision Optimization experiments

5 March 2025

When building Decision Optimization models in the experiment UI, you can now use code snippets for Python DOcplex or OPL models. Using code snippets can make model building faster, as you can add and edit code without having to enter all the lines of code from scratch.

For more information, see Code snippets for building models.

Removal of Runtime 23.1

27 February 2025

Support for IBM Runtime 23.1 in watsonx.ai Runtime and watsonx.ai Studio will be removed on 17 April 2025. To ensure a seamless experience and to leverage the latest features and improvements, switch to IBM Runtime 24.1.

• For information about changing environments, see Changing notebook environments.
• For details on deployment frameworks, see Managing frameworks and software specifications.

Deprecation of project-lib library

24 February 2025

The project-lib library is deprecated. Starting with Runtime 25.1, the library will not be included in any new runtime version. Although existing runtime versions through 24.1 will continue to include the deprecated library, consider rewriting your code to use the ibm-watson-studio-lib library.

For information on how to migrate your code, see:

• Migrating from project-lib for Python to ibm-watson-studio-lib
• Migrating from project-lib for R to ibm-watson-studio-lib

Relationship explorer to visualize your metadata

21 February 2025

Relationship explorer is now available to help better understand your data. This new feature helps you to visualize, explore and govern your metadata. Discover how your governance artifacts and data assets relate with each other in a single view.

For more information, see Relationships.

Assign aliases for more complete lineage in Manta Data Lineage

21 February 2025

In complex data environments that connect multiple systems and technologies, lineage might appear incomplete due to missing system connections. You can now assign aliases to systems to bridge these gaps and generate a more complete and accurate cross-system lineage.

For more information, see Configuring alias assignments.

Deploy models converted from scikit-learn and XGBoost to ONNX format

13 February 2025

You can now deploy machine learning and generative AI models that are converted from scikit-learn and XGBoost to ONNX format and use the endpoint for inferencing. For more information, see Deploying models coverted to ONNX format.

IBM watsonx.data intelligence and Manta Data Lineage are now also available in the Toronto region

14 February 2025

IBM watsonx.data intelligence and Manta Data Lineage are now also available in the Toronto data center. You can select Toronto as your preferred region when you sign up.

For more information about product features that are available in the Toronto region, see Regional availability for services and features.

Deploy models converted from scikit-learn and XGBoost to ONNX format

13 February 2025

You can now deploy machine learning and generative AI models that are converted from scikit-learn and XGBoost to ONNX format and use the endpoint for inferencing. For more information, see Deploying models coverted to ONNX format.

Updated SPSS Modeler tutorial videos

11 February 2025

Watch and learn about SPSS Modeler by viewing the updated videos in the SPSS Modeler tutorials.

Deprecation of the Recommended assets tab (IBM watsonx.data intelligence)

4 February 2025

The Recommended tab on the Asset page with assets recommended to you based on properties common to the assets that you viewed, created, and added to projects is deprecated and will be removed in March 2025.

If you have questions or concerns that are related to the deprecation, you can open a support ticket.

Default Inventory replaces Platform Asset Catalog in watsonx.governance

3 February 2025

A default inventory is now available to store watsonx.governance artifacts including AI use cases, third-party models, attachments, and reports. The default inventory replaces any previous dependency on Platform access catalog or IBM Knowledge Catalog for storing governance artifacts.

Manta Data Lineage is now also available in the Sydney region

21 January 2025

Manta Data Lineage is now also available in the Sydney data center. You can select Sydney as your preferred region when you sign up.

For more information about product features that are available in the Sydney region, see Regional availability for services and features.

Deploy models converted from CatBoost and LightGBM to ONNX format

15 January 2025

You can now deploy machine learning and generative AI models that are converted from CatBoost and LightGBM to ONNX format and use the endpoint for inferencing. These models can also be adapted to dynamic axes. For more information, see Deploying models coverted to ONNX format.

New Evaluation Studio tutorial and video

13 Jan 2025

Try the new Evaluation Studio tutorial and video to help you learn how to evaluate and compare the performance of your generative AI assets.

Deprecation of data location and sovereignty rules for Data Privacy

13 January 2025

The data location and sovereignty rules are experimental features that provide attributes-based access control of data assets based on their location or sovereignty. These experimental features are deprecating and might be removed in March 2025. For details, see Data location rules (experimental).

If you have questions or concerns that are related to the deprecation, you can open a support ticket.