What's new and changed in watsonx.governance

IBM watsonx.governance updates can include new features, security updates, and fixes. Releases are listed in reverse chronological order so that the latest release is at the beginning of the topic.

You can see a list of the new features for the platform and all of the services at What's new in IBM Cloud Pak for Data.

Installing or upgrading watsonx.governance

Ready to install or upgrade watsonx.governance?

  • To install watsonx.governance along with the other Cloud Pak for Data services, see Installing Cloud Pak for Data.
  • To upgrade watsonx.governance along with the other Cloud Pak for Data services, see Upgrading Cloud Pak for Data.
  • To install or upgrade watsonx.governance independently, see watsonx.governance.
    Remember: All of the Cloud Pak for Data components associated with an instance of Cloud Pak for Data must be installed at the same version.

Cloud Pak for Data Version 5.0.3

A new version of watsonx.governance was released in September 2024 with Cloud Pak for Data 5.0.3.

Operand version: 2.0.3

This release includes the following changes:

New features
This release of watsonx.governance includes the following features:
Associate workspaces with AI use cases

The flow for creating an AI use case now more closely aligns with the AI lifecycle. After you define the essentials of an AI use case, you must now associate workspaces to organize assets so they align with the phases of your AI solution. For example, you can associate a project or space for assets in the Development or Validation phases, and associate a space for assets in the Operation phase.

If you created AI use cases in previous versions of watsonx.governance, follow the migration instructions to associate workspaces with the use cases.

For details, see Associating workspaces with an AI.

Select test data for prompt template evaluations from projects or spaces

When you evaluate prompt templates, you can now choose assets from projects or deployment spaces to select test data for evaluations.

For details, see Evaluating prompt templates in spaces or Evaluating prompt templates in projects.

Updates
The following updates were introduced in this release:
New drift v2 evaluation metric
You can now calculate the embedding drift metric to detect the percentage of records that are outliers when compared to the baseline data. For more information, see Drift v2 evaluations
Calculate retrieval quality metrics with Python SDK
Retrieval quality metrics measure how the retrieval system identifies and ranks relevant contexts. You can now calculate the following retrieval quality metrics with the Python SDK:
  • Context relevance
  • Retrieval precision
  • Average precision
  • Reciprocal rank
  • Hit rate
  • Normalized Discounted Cumulative Gain

For more information, see Metrics computation with Python SDK.

Use LLM-as-a-judge to generate RAG metrics with Python SDK
You can now use LLM-as-a-judge models to measure the performance of RAG tasks by calculating answer quality and retrieval quality metrics with the Python SDK. For more information, see Metrics computation with Python SDK.
Configure fairness evaluations in watsonx.governance
You can now configure fairness evaluations in watsonx.governance to determine whether LLM models produce biased outcomes. For more information, see Fairness evaluations.
New quality evaluation metrics for regression models
You can now configure the following new quality metrics for regression models:
  • Mean absolute percentage error
  • Symmetric mean absolute percentage error
  • Pearson correlation coefficient
  • Spearman correlation coefficient

For more information, see Quality evaluations.

New azure studio deployments
You can now use Azure designer container instance endpoints for Azure Studio deployments.
View RAG evaluation results with root cause analysis
You can now analyze prompt template evaluation results for RAG tasks with root cause analysis.
Security issues fixed in this release
The following security issues were fixed in this release:
AI Factsheets

CVE-2024-0450, CVE-2024-20697, CVE-2024-20696, CVE-2024-21890, CVE-2024-21896, CVE-2024-24806, CVE-2024-28182, CVE-2024-2961, CVE-2024-33599, CVE-2024-33600, CVE-2024-33601, CVE-2024-33602

CVE-2023-39331, CVE-2023-6597

CVE-2022-0391, CVE-2022-25881

OpenPages
This release includes bug fixes and security fixes. For more information, see OpenPages with Watson 9.0 Fix List.

Cloud Pak for Data Version 5.0.1

A new version of watsonx.governance was released in July 2024 with Cloud Pak for Data 5.0.1.

Operand version: 2.0.1

This release includes the following changes:

New features
This release of watsonx.governance includes the following features:
Associate attachments with AI lifecycle phases
If you are an inventory owner, you can now add attachment placeholders to specific lifecycle phases for AI Factsheets. For example, you can add placeholders to prompt users to:
  • Attach approval documents to the Develop phase.
  • Attach details about validation data to the Validate phase.
  • Attach usage details to the Operate phase.
Attachments display in the designated sections of AI Factsheets, creating a more organized audit trail. For details, see Managing AI use cases.
Updates
The following updates were introduced in this release:
Calculate new RAG metrics with Python SDK
You can now use the Watson OpenScale Python SDK to calculate the following new metrics that can evaluate how well your LLM performs RAG tasks:
  • Context relevance
  • Faithfulness
  • Answer relevance
  • Answer similarity

For more information see, Metrics computation with Python SDK.

Select metrics to configure prompt template evaluations
You can now select the metrics that you want to use evaluate your prompt templates in watsonx.governance projects or spaces when you configure evaluation settings.
Security issues fixed in this release
The following security issues were fixed in this release for each component:
AI Factsheets

CVE-2024-0450, CVE-2024-20697, CVE-2024-20696, CVE-2024-21890, CVE-2024-21896, CVE-2024-24806, CVE-2024-28182, CVE-2024-2961, CVE-2024-33599, CVE-2024-33600, CVE-2024-33601, CVE-2024-33602

CVE-2023-39331, CVE-2023-6597

CVE-2022-0391, CVE-2022-25881

OpenPages
This release includes bug fixes and security fixes. For more information, see OpenPages with Watson 9.0 Fix List.
Watson OpenScale

CVE-2024-35195

CVE-2024-36114

CVE-2023-45288

Cloud Pak for Data Version 5.0.0

A new version of watsonx.governance was released in June 2024 with Cloud Pak for Data 5.0.0.

Operand version: 2.0.0

This release includes the following changes:

New features
This release of watsonx.governance includes the following features:
Assess use cases for EU AI Act applicability
By using the new EU AI Act applicability assessment, you can complete a simple questionnaire to assess your AI use cases and determine whether they are within the scope of the EU AI Act. The assessment can also help you to identify the risk category that your use cases align to: prohibited, high, limited, or minimal.
For more information, see Applicability Assessment in Solution components in Governance console.
Create detached deployments for governing prompts for externally hosted large language models (LLMs)
A detached prompt template is a new asset for evaluating a prompt template for an LLM that is hosted by a third-party provider, such as Google Vertex AI, Azure OpenAI, or AWS Bedrock. The inferencing that generates the output for the prompt template is done on the remote model, but you can evaluate the prompt template output by using watsonx.governance metrics. You can also track the detached deployment and detached prompt template in an AI use case as part of your governance solution.
For more information, see:
New metrics for evaluating prompt templates
When you evaluate prompt templates in your watsonx.governance deployment spaces or projects, you can now run generative AI quality evaluations to measure how well your model performs retrieval-augmented generation (RAG) tasks with the following new metrics:
  • Faithfulness
  • Answer relevance
  • Unsuccessful requests
Results from these new evaluations are captured in factsheets in AI use cases.

For more information, see Generative AI quality evaluations.

Updates
The following updates were introduced in this release:
Calculate RAG metrics with Python SDK
You can now use the Watson OpenScale Python SDK to calculate metrics that can evaluate how well your LLM performs RAG tasks. These metrics include:
  • Content analysis
  • Keywords inclusion
  • Question robustness
Generate drift v2 evaluation metrics for additional data types
When you enable drift v2 evaluations, you can now generate the prediction drift, output drift, and input metadata drift metrics to measure the performance of unstructured text and unstructured image models. You can also generate the prediction drift and input metadata drift metrics for structured models.
Security issues fixed in this release
The following security issues were fixed in this release:

CVE-2024-0727, CVE-2024-22365, CVE-2024-2398, CVE-2024-24786, CVE-2024-25062, CVE-2024-2511, CVE-2024-25260, CVE-2024-28834, CVE-2024-28835, CVE-2024-2961, CVE-2024-3154, CVE-2024-3177, CVE-2024-33599, CVE-2024-33600, CVE-2024-33601, CVE-2024-33602, CVE-2024-34064, CVE-2024-3651

CVE-2023-2953, CVE-2023-2975, CVE-2023-3446, CVE-2023-3817, CVE-2023-45288, CVE-2023-45322, CVE-2023-5678, CVE-2023-6129, CVE-2023-6237, CVE-2023-7008

CVE-2022-27943, CVE-2022-41409, CVE-2022-48554

CVE-2021-33294, CVE-2021-3997

CVE-2020-12413, CVE-2020-35512, CVE-2020-8565

CVE-2019-11250

CVE-2018-20839