The landscape of business intelligence is moving at a relentless pace, and the latest Power BI news reveals a platform that is fundamentally transforming. As of April 2026, the focus has shifted from mere data visualization to an AI-first, developer-friendly ecosystem. This transition involves not only the introduction of sophisticated tools but also the phased retirement of legacy features that have been staples for a decade. Understanding these shifts is essential for maintaining robust reporting environments and leveraging the next generation of analytics.

The rise of Copilot and the sunset of legacy natural language tools

One of the most significant themes in recent Power BI news is the definitive transition from the legacy Q&A visual to the more advanced Copilot experience. For years, Q&A allowed users to ask questions in natural language, but its limitations in understanding complex context and semantic nuances were often apparent. Microsoft has confirmed that Q&A is on a deprecation path, with a full retirement scheduled for December 2026.

In its place, Power BI Copilot has taken center stage. Unlike its predecessor, Copilot utilizes large language models to provide deeper insights, generate entire report pages, and even suggest DAX measures. A major update recently made the standalone Copilot experience available directly on the Power BI Home page. This change is designed to reduce friction, allowing users to start querying their data as soon as they log in, without having to navigate to a specific report first.

The "Approved for Copilot" governance model

To manage this AI expansion, a new governance setting originally known as "Prepped for AI" has been rebranded as "Approved for Copilot." This is a critical development for data architects. When a semantic model is marked as "Approved for Copilot," it gains several advantages:

  • It bypasses certain "friction treatments" in the standalone Copilot agent, leading to faster responses.
  • It is prioritized in search results within the Copilot interface.
  • Admins can restrict the standalone Copilot to only use these approved models, ensuring that users are interacting with verified, high-quality data.

This move suggests a shift toward more disciplined data curation. Instead of letting AI roam across every messy dataset in a tenant, organizations are encouraged to certify their models specifically for AI consumption.

PBIR: A new era for report development and version control

For developers, perhaps the most impactful Power BI news this year is the transition to the Power BI Enhanced Report Format (PBIR) as the default for Power BI Project (PBIP) files. For a long time, the .pbix format was a "black box," making it difficult to track changes in version control systems like Git or to perform bulk edits.

PBIR changes the game by using a folder-based structure with human-readable JSON files. This allows multiple developers to work on the same report simultaneously without constant merge conflicts. The latest updates have made PBIR the standard, signaling that the platform is moving toward a more mature, code-centric development lifecycle. This is particularly relevant for large-scale enterprise deployments where change management and audit trails are mandatory.

The impact on CI/CD pipelines

With PBIR becoming the norm, the integration with Fabric and Git is now more seamless. The ability to see exactly what changed in a visual's properties or a filter's configuration within a Git diff is a massive productivity boost. It also opens the door for third-party tools and scripts to generate or modify reports programmatically, a capability that was previously limited and prone to breaking.

Critical deadlines: Feature retirements in April and May 2026

Staying current with Power BI news requires a close eye on the retirement calendar. We have reached several critical milestones this month that demand immediate attention from report authors and server administrators.

SSRS Report Viewer Web Part support ends

As of April 13, 2026, support for the SQL Server Reporting Services (SSRS) Report Viewer web part for SharePoint has officially ended. While the component may still be available for download in some legacy repositories, it will no longer receive security updates or bug fixes. Organizations still relying on embedding on-premises paginated reports into SharePoint Server pages should consider migrating to Power BI paginated reports or utilizing the modern Power BI embedding solutions. Ignoring this change could lead to security vulnerabilities or broken report links in corporate portals.

Imminent deprecation of R and Python visuals

Looking forward to next month, May 2026 will mark the end of support for R and Python visuals in "Embed for your customers" (App owns data) and "Publish to web" scenarios. This is a vital update for developers who have built custom analytics products using these languages. After the May deadline, these charts will likely display as blank spaces in embedded applications.

While R and Python remain supported for internal organizational use and secure embedding, their removal from public-facing and customer-facing scenarios reflects a strategic pivot toward using more standardized, high-performance visuals or custom visuals built with the Power BI SDK. If your application relies on these visuals, the current window for migration is closing rapidly.

Reporting enhancements: Azure Maps and Modern Tooltips

In the realm of visualization, the push to replace legacy Bing Maps with Azure Maps is nearly complete. Recent updates have introduced "Markers" to the Azure Maps visual, allowing for more detailed and data-driven location storytelling without needing complex GIS tools. The Bing Maps icon is being removed from the default visualization pane, so any remaining reports using the old mapping technology should be updated to ensure continued functionality and better performance.

Modern Visual Tooltips are now Generally Available

Another subtle but powerful change is the General Availability (GA) of modern visual tooltips. These tooltips provide a cleaner look and feel, supporting better interaction and drill-through capabilities. They also align more closely with the overall Fabric UI, providing a consistent experience as users move between Power BI, Data Factory, and Real-Time Intelligence modules.

Modeling breakthroughs: Composite models on Direct Lake

One of the most technically impressive pieces of Power BI news involves the public preview of composite semantic models that combine Direct Lake tables with Import tables. Historically, users had to choose between the lightning-fast performance of Import mode and the real-time, zero-copy benefits of Direct Lake (associated with Fabric Lakehouses).

Now, the ability to mix these modes in a single model offers unprecedented flexibility. For example, a developer can keep massive historical sales data in Direct Lake to avoid long refresh times, while importing a small, frequently changing Excel file for local targets. This "hybrid" approach is likely to become the standard architecture for complex enterprise models throughout 2026.

Recap from FabCon Atlanta

March 2026 saw the gathering of the global data community at FabCon Atlanta. The news coming out of that event centered on the concept of "The Unified Data Platform." Microsoft has doubled down on making Power BI the primary consumption layer for all Fabric workloads.

Key takeaways from the conference sessions included:

  • Semantic Model Version History: This feature is now GA, allowing users to restore previous versions of their models directly from the Power BI service—a safety net that has been high on the community's wishlist.
  • DAX User-Defined Functions (UDFs): In preview, this allows for the creation of reusable logic within a model, reducing the need to copy-paste complex DAX formulas across multiple measures.
  • Task Flows in Power BI: The integration of translytical task flows, which are now enabled by default, helps report creators visualize the end-to-end journey of data from source to report.

Practical advice for the 2026 landscape

Given the volume of changes, it is important to prioritize actions based on risk and opportunity. Here are a few suggested focus areas for the coming months:

  1. Audit for R/Python Visuals: If you provide analytics to external customers via embedded reports, verify immediately if R or Python visuals are in use. These need to be replaced before the May 2026 cutoff.
  2. Evaluate PBIR Migration: For teams using Git for version control, transitioning to the PBIR format for new projects will likely yield significant improvements in development speed and reliability.
  3. Governance for Copilot: Start designating specific semantic models as "Approved for Copilot." This not only prepares your organization for the AI shift but also acts as a quality control mechanism to prevent AI-generated hallucinations based on unverified data.
  4. Review Mapping Visuals: If any older reports still use Bing Maps, transition them to Azure Maps. The feature parity is now complete, and the older visual is officially being phased out of the default UI.

The path forward

Power BI news in 2026 paints a picture of a platform that is shedding its legacy skin to become something more integrated and intelligent. The deprecation of familiar tools like Q&A and certain embedding options might feel disruptive, but they are necessary steps to make room for a more scalable, AI-integrated future. By staying informed of these deadlines and new capabilities, data professionals can ensure their reporting ecosystems remain efficient, secure, and ready for the next wave of innovation.