Skip to main content
UBSW.EU

AI governance · Rank 03

IBM watsonx.governance

AI use-case inventory, EU AI Act applicability assessment, governance workflows and model evaluation and monitoring.

100 points across 25 criteria · public vendor evidence

Profile overview

At a glance

Category

AI governance

Evidence

Public vendor pages

Not found

1 of 25 criteria

Reviewed

13 November 2026

25 criteria · 100 points

Criterion-by-criterion

Each criterion is worth four points: 4 when the reviewed pages clearly document it, 2 for narrower or partial evidence, and 0 when the vendor says it is absent or no supporting statement was found. “Not found” is marked separately from an explicit absence.

IInventory and scope16 / 20
I1Central AI system inventoryA searchable register of AI systems or use cases.Documented4/4
I2Third-party AI includedVendor, embedded or externally supplied AI can be recorded.Partially documented2/4
I3Discovery beyond manual entryA documented way to discover AI in the organisation or connect discovery sources.Partially documented2/4
I4Purpose and ownership recordedA use case can be tied to its purpose and accountable people or teams.Documented4/4
I5Generative AI and agents coveredThe published product scope explicitly includes generative AI, agents or both.Documented4/4
RRisk and regulation16 / 20
R1Risk assessment and classificationA documented workflow assesses or classifies AI use cases by risk.Documented4/4
R2EU AI Act mappingA product workflow, template or controls are expressly mapped to the EU AI Act.Documented4/4
R3More than one governance frameworkThe product maps controls to additional frameworks such as ISO/IEC 42001 or NIST AI RMF.Documented4/4
R4Impact assessment workflowA documented impact or fundamental-rights assessment is available.Not found in reviewed pages0/4
R5Responsibilities and stakeholdersThe workflow assigns or routes review across accountable functions or stakeholders.Documented4/4
CControls and workflow18 / 20
W1Structured intakeNew AI use cases enter through a documented request or intake workflow.Documented4/4
W2Approval gatesReviews or approvals can be required before deployment or material changes.Documented4/4
W3Role-based governanceThe product documents differentiated roles or access for governance work.Partially documented2/4
W4Human oversight recordedHuman oversight responsibilities or checkpoints can be documented in the workflow.Documented4/4
W5Policies become controlsPolicies or regulatory requirements map to trackable controls or actions.Documented4/4
EEvidence and records18 / 20
E1Technical evidence and documentationThe product records or manages technical documentation and supporting evidence.Documented4/4
E2System relationships and lineageLinks between systems, models, data, vendors or dependencies can be represented.Partially documented2/4
E3Traceable review historyActions, changes or reviews have a documented history or audit trail.Documented4/4
E4Governance reportingA governance report, factsheet or evidence export is documented.Documented4/4
E5Evidence tied to controlsEvidence, sign-offs or validation results can be attached to governance requirements.Documented4/4
LLifecycle assurance16 / 20
L1Post-deployment monitoringMonitoring continues after deployment, beyond a one-time approval.Documented4/4
L2Performance and responsible-AI testingTesting or monitoring includes quality, fairness, bias, robustness or drift.Documented4/4
L3Incident or issue follow-upDetected issues can trigger a documented response, workflow or remediation.Partially documented2/4
L4Runtime controlsThe product documents guardrails or controls that can act during AI use.Partially documented2/4
L5Periodic review and changeThe product documents reassessment, review intervals or lifecycle change handling.Documented4/4

Evidence trail

Vendor pages reviewed

Only public vendor product pages and documentation were reviewed. Sales demos, contracts and security packs were not examined. Features can vary by plan, region and configuration.

Method

How this score is produced

The score measures the coverage of a vendor’s public evidence against a common feature rubric. It does not certify that the product meets the AI Act or that a workflow works in practice.

4 points

The reviewed vendor pages clearly document the capability.

2 points

Public evidence describes a narrower or partial version of the criterion.

0 points

The vendor explicitly says it is absent, or no supporting evidence was found. “Not found” is flagged separately.

Category scope, inclusion rules and sources are documented on AI governance.