Release date:
This release introduces Agentic Governance, a new family of autonomous agents starting with the Duplicate Content Detection Agent for Power BI, which continuously monitors your BI estate and surfaces duplicate reports for you to act on.
Power BI also gains a new Security & Compliance quality gate for BI Asset Policies and deeper Capacity Planning insights with new global KPIs.
Tableau sees a significant expansion of Regression Test (Cross Environment) with new non-visual comparison checks, the arrival of Priority Monitoring for Tableau dashboards, and AI-powered recommendations for resolving failed Render Duration policies.
As always, you can find important links just below.
📖Full Documentation 💡Knowledge Base 📞Customer Support
This release marks the start of something new for Wiiisdom Cloud Platform: Agentic Governance. Rather than waiting for someone to run a check or read a report, governance agents work continuously in the background, monitoring your BI estate and surfacing what matters, so your team can spend less time hunting for problems and more time fixing them.
The first agent to join the family is the Duplicate Content Detection Agent for Power BI. It runs autonomously against your tenant, identifying clusters of reports that are likely duplicates of one another, based on shared semantic models, structural similarity between pages and visuals, and usage patterns, so redundant content stops quietly accumulating across workspaces.
Getting an agent running takes a couple of clicks. From the BI Platforms page, a new "Agentic Governance" configuration screen lets you enable the Duplicate Content Detection Agent on a given platform, assign it a runner, and see at a glance whether it's active.
Once running, the agent's discoveries land on a new Findings page. Each finding shows the cluster of likely-duplicate reports, the agent's reasoning for flagging them (usage numbers, structural overlap, shared data sources), a recommended action such as consolidating into a single surviving report, and an impact rating so you know what to prioritise first. From there, you can dismiss a finding, snooze it, notify the report owners directly, or mark it resolved.
Every action taken on a finding, and every finding the agent has ever surfaced, is preserved in a new Audit view, giving you a full, traceable history of what the agent found and what your team did about it.
This is just the first of many governance agents on the roadmap. Duplicate Content Detection is arriving for Power BI today, with equivalent agents for Tableau planned for a future release.
BI Asset Policies for Power BI now includes a new quality gate: Security & Compliance. It groups governance checks specifically around data protection and report safety, alongside the existing Usability and Performance categories.
The new gate includes rules for Export Data Strategy (controlling what level of data users can export from a report), Row Level Security (flagging whether a semantic model has RLS configured), and Custom Visual usage (flagging reports that rely on custom visuals, which can affect performance).
Each rule can be configured with its expected value directly from the BI Asset Policies configuration page.
The Power BI Capacity Planning view now includes a set of global KPIs at the top of the page, giving you an immediate summary of the capacity's recent usage, including average and peak consumption and the workspace consuming the most.
BI Asset Policies for Power BI is now fully supported for tenants connected through the Wiiisdom Cloud Platform Gateway, with the Render Duration rule joining the other policy checks already available in that mode.
This means Power BI customers running in a restricted network configuration can now use the complete BI Asset Policies feature set, matching what's available in the standard cloud connection, with no functional gaps between the two.
The Open Report step is now fully supported in pipelines for Power BI tenants connected through the Wiiisdom Cloud Platform Gateway, matching its behaviour in the standard cloud connection.
With this addition, every pipeline step now runs through the Wiiisdom Cloud Platform Gateway with one exception: Simulation Journey. We're closing in fast on full pipeline parity between the standard cloud connection and Gateway-connected tenants.
We've significantly expanded what you can test with the Regression Test (Cross Environment) step for Tableau. Alongside the existing image-based visual comparison, the step now runs non-visual regression checks directly in the cloud platform, opening up a whole new layer of comparisons for catching changes between workbook versions.
This release introduces the first set of checks: comparing summary and full columns, comparing summary data, and checking whether filters exist and what type they are.
Additional checks, covering filter values, filter domain values, and parameter data types and current values, are planned for upcoming releases.
These checks bring the Regression Test (Cross Environment) step closer to parity with the non-visual regression checks already available in Wiiisdom for Tableau Desktop.
Priority Monitoring, previously available for Power BI, now extends to Tableau. The Priority Monitoring page surfaces the Tableau dashboards that need attention right now, ranked by a priority score built from recent usage and performance trends, alongside a forecasted score for the next business day.
This is the first step of Priority Monitoring's rollout to Tableau, with further coverage and refinements to follow in upcoming releases.
Tableau functional testing, previously only available in Wiiisdom for Tableau Desktop, is now available directly in Wiiisdom Cloud Platform through a new pipeline step: Simulation Journey.
Rather than manually building each test step one by one, you can now describe what you want to test in plain language, and an AI assistant generates the corresponding steps for you, including opening views, applying filters and parameters, and asserting expected data, images, or worksheet content.
Every step generated this way is clearly marked as AI-generated, so you can review it before relying on it, and any step can be edited or removed if it doesn't match what you intended.
When a Tableau workbook fails its Render Duration policy during a Workbook Policy Enforcement step, the execution report now offers a "Show recommendations" button on the failed rule. Clicking it surfaces AI-generated, contextual guidance on what's driving the slow render time and what to change in the workbook to bring it back under the allowed limit.
Recommendations are generated at the moment the pipeline runs and are ranked by importance, so the most impactful changes are the ones you see first.
The Governance Digest email, previously available for Power BI, now extends to Tableau, giving you one unified governance inbox across both technologies. Each digest clearly indicates whether the underlying platform is Tableau Cloud or Tableau Server, making it easy to track governance across both if you run them in parallel.
The Tableau section surfaces key insight metrics including number of workbooks, average views and opening times per workbook, datasource count, average refresh time, and datasource size, along with call-outs for workbooks that haven't been viewed and datasources that are refreshed without any usage.
Column Statistics pipeline steps can now be created and updated for Tableau through the public API and through Wiiisdom's MCP integration, in addition to Power BI.
Power BI platforms can now be connected as Admin-Managed, an alternative to the existing delegated connection model. An Admin-Managed platform requests a reduced set of Microsoft permissions at connection time and requires a single Runner (a Microsoft Entra ID Service Principal) through which all pipeline executions are routed.
Before any execution, Wiiisdom Cloud Platform checks that both the user and the Runner have access to the target workspace, so actions can never run outside what either identity is actually permitted to touch. The same check applies at configuration time: when building a pipeline or a BI Context, only workspaces accessible to both identities appear in the selector.
Execution reports reflect this model clearly, showing the Runner's identity for any step it executed, so it's always visible which identity performed which action.
Setting up a Power BI Runner for Predictive Monitoring now includes a built-in diagnostic that checks your configuration end-to-end before you rely on it. It verifies the runner can connect, walks through the required Power BI tenant settings step by step, and confirms access to Fabric capacity metrics.
Each check reports a clear result, success, warning, or failure, with actionable guidance when something needs attention, such as a tenant setting to enable or a permission to grant. This makes it much faster to catch and fix configuration issues upfront, rather than discovering them partway through a monitoring run.
A user's BI platform credentials, including Power BI refresh tokens and Tableau personal access tokens, are now automatically invalidated whenever that user's email address is changed. Affected BI contexts will need to be reconfigured before pipelines using them can run again.
We have upgraded the embedded Keycloak identity provider to version 26.7.2 to resolve CVE-2026-18963, a vulnerability in the credential reset flow.
▪ Site tree indentation missing when expanding projects
Medium impact · WOFT-3447
Description: In the site tree used by the Row Count and Simulation Journey steps, expanding a project did not add indentation to its contents, making it difficult to tell which objects belonged to which parent project.
▪ "Deploy entire workspace" toggle not adding reports correctly
Medium impact · WIIIS-2428
Description: In the Deploy Content Cross Env step, enabling the "Deploy entire workspace" toggle did not add all reports to the selected content panel, left the Finish button disabled, and could cause previously configured report parameters to be lost.
▪ BI Policies Enforcement errors with Calculation Groups
Medium impact · WIIIS-2418
Description: The BI Policies Enforcement step returned errors for rules such as Connectivity Mode, Row Level Security, and Number of Rows when the linked semantic model contained a Calculation Group table.
▪ Column Statistics step failing on DAX queries with special characters
Medium impact · WIIIS-2417
Description: The Column Statistics step in DAX mode failed when the DAX query included a column name containing parentheses.
▪ Webhook notifications not sent due to conflict with email post-process
Low impact · WCPC-383
Description: In some pipeline executions, webhook notifications could fail to send because of an incorrect interaction with the email notification post-process.
