Microsoft released the Power BI February 2026 feature update last week, and there are several changes worth paying attention to if you run reporting or data tools in your business.
I’ve been working through the update this week and wanted to share what stood out.
The February release touches several parts of the platform, from how you filter and paste data into reports through to how Copilot reads and summarises your apps.
Below I’ll walk through the Input Slicer, Conversational Copilot, card visual updates, new DAX functions, and some important deprecation notices with deadlines coming mid-2026.
If you use Power BI day-to-day, there are a few things here worth acting on.
Ditch the Slicer Struggle
Picture this: you’re in a board meeting and the sales director emails a list of 50 priority clients from Queensland. Normally, you’d spend 20 minutes manually selecting them in Power BI slicers, one by one, while the room waits. That’s a frustrating and avoidable problem.
The new Input Slicer, now generally available (GA), addresses this directly. You type or paste free-text values like customer IDs, product codes, or postcodes, and it filters using “contains any” or “starts with” logic.
Figure: Input slicer in Power BI report showing the filter options menu (e.g., contains, starts with).
It also works as a task flow input for write-backs, which means it can feed data directly into Microsoft Fabric workflows. (Fabric is Microsoft’s unified data and analytics platform, bringing together data storage, reporting, and AI in one place.)
This means instant regional filtering without building custom visuals.
Paste into Slicers takes it one step further. Copy a list from Excel, an email, or a Teams chat and drop it straight into the slicer. No retyping required.
Figure: Paste a list of values into a slicer using the slicer header menu (… > Paste).
Copilot Gets Conversational
Fabric Copilot capacities are now enabled by default, tenant-wide, from 8 February 2026. No configuration required; your teams get AI-assisted reporting out of the box.
The headline feature here is Conversational Copilot in Power BI apps.
Rather than requiring you to pre-select a report, Copilot scans your entire app, finds the relevant data, and returns summaries or visuals based on plain-language questions like “Show QLD revenue trends vs last year.
Figure: Item selection is no longer required. Copilot automatically chooses the most relevant item.
If the data is ambiguous, it asks a clarifying question before responding, which keeps results accurate rather than just fast.
Figure: When more than one relevant item is found, app Copilot will ask a brief clarifying question before choosing the best item to answer from.
I’ve seen this work well for hybrid Australian teams. Executives on the Gold Coast can query “Melbourne client churn” from their phone and get a clear visual without waiting on a BI developer. The Power BI app summaries also now auto-generate overviews that can flag spending spikes or unusual patterns in your data.
For businesses I work with across manufacturing and services, this is also genuinely useful. Non-technical managers get the same quality of reporting that used to sit only with senior analysts, and it reduces pressure on already-stretched IT teams.
Cards That Command Attention
The Card visual has received a significant update. It now defaults to 10 callout items, covering KPIs like revenue, margins, and headcount, with interactive category filtering built in.
Figure: New Card visual showing multiple callouts (default increased to 10).
Click a category and it drills down, with no extra visuals needed.
Cross-platform rendering has also been improved, so your Brisbane dashboard displays correctly on a MacBook in a Melbourne meeting room.
Figure: Default report fonts (for example, Segoe UI) now render correctly when viewing Power BI reports on non-Windows devices.
Tooltips are cleaner too. Single-value cards show trend indicators with directional icons, while multi-row cards can display category images and top-N records (a filtered list of your highest-priority items).
As an example, a card showing “AU Sales: $2.4M (+12%)” with a green up-arrow could then filter to “QLD only” to reveal which regions are performing.
For our clients focused on Microsoft 365 integration, these card visuals pair well with Copilot. A card showing Fabric lakehouse metrics (your centralised data storage layer) can trigger a Copilot query like “Why did costs rise?” and bring the answer back into the same view.
Good News for DAX Developers
If your team builds custom data models, the new DAX functions NAMEOF and TABLEOF will save a lot of rework. DAX (Data Analysis Expressions) is the formula language used in Power BI to build calculations and data models.
Previously, if a column or table was renamed, any formulas referencing it would break. With NAMEOF and TABLEOF, formulas automatically adapt when names change. You write the reference once and it survives renames.
TABLEOF works the same way for full tables, which is particularly useful for user-defined functions (custom calculations you build and reuse) in Fabric semantic models. A semantic model is the organised data layer that Power BI reports read from.
This also benefits businesses connecting .NET or Python applications via the new ODBC driver preview.
Australian businesses running hybrid data environments, with both on-premises SQL databases and Azure cloud data, can avoid report refresh failures when their data schema changes.
We use this approach in Fabric migration projects to keep BI assets working reliably as businesses scale.
Sharing Reports Just Got Better
PBIR (Power BI Report format) is now the default export format. This means reports shared between team members retain full interactivity, rather than being flattened into PDFs that can’t be filtered or explored.
Azure Maps visuals also received a performance update this month, with faster rendering when working with pie chart overlays. Azure Maps now supports two new endpoint regions, Korea and Brazil, giving businesses more flexibility on where their data is processed.
Copilot Governance Controls
The Fabric Copilot capacity tenant setting is now enabled by default from 8 February 2026.
Administrators can manage which capacities are designated for Copilot use, giving IT teams control over how and where AI features are rolled out across the business.
Action Required: Legacy Excel and CSV Import Being Retired
The legacy Excel and CSV import experience in Power BI Service is being retired. Entry points will be removed on 31 May 2026, and semantic models created using that legacy import will stop refreshing on 31 July 2026.
If your team still uses this approach, consider a move to Fabric’s OneLake.
OneLake is Microsoft’s unified storage layer that connects all your data in one place without duplicating it.
What This Means for Your Business
These updates are about reducing the manual work that sits between your data and your decisions.
Input slicers and paste functions cut the time spent setting up reports, freeing teams to focus on what the data actually means. Conversational Copilot means business leaders can get answers directly from their data without always needing a developer in the loop.
The use cases I keep seeing across our client base follow a similar pattern:
Energy businesses paste rig IDs into slicers and surface downtime patterns in minutes, while Copilot returns a plain-language summary of the top cost drivers across site operations
Not-for-profit teams use card callouts to track grant funding and donor activity by postcode, filtered on the fly without needing a BI developer
Manufacturing businesses monitor cost trends across multiple sites and use Copilot to flag anomalies before they become problems
For Australian businesses, there’s an added layer of confidence here. Fabric runs on local Australian data centres, which means data stays onshore, performance is consistent, and businesses in regulated sectors like energy, healthcare, and government can meet local data handling requirements without additional configuration.
Your Next Steps
If you want to start putting these features to work, here’s where I’d begin:
Check that tenant-wide Copilot is active in your Fabric workspace settings (it should be on by default from 8 February 2026).
Test the Input Slicer on a live sales report by pasting in a real client list.
Build a card dashboard with five to seven callouts for your most important KPIs.
Work with your admin team to approve the right semantic models before enabling Copilot more broadly.
Review any semantic models created using the legacy Excel or CSV import experience in Power BI Service and plan to move these to OneLake shortcuts before 31 May 2026.
If you’d like to talk through how these updates apply to your business, feel free to reach out via LinkedIn or click the below to book a discovery session with one of our consultants.
About the Author
Carlos Garcia is the Founder and Managing Director of CG TECH, where he leads enterprise digital transformation projects across Australia.
With deep experience in business process automation, Microsoft 365, and AI-powered workplace solutions, Carlos has helped businesses in government, healthcare, and enterprise sectors streamline workflows and improve efficiency.
He holds Microsoft certifications in Power Platform and Azure and regularly shares practical guidance on Copilot readiness, data strategy, and AI adoption.
Microsoft released the Power BI February 2026 feature update last week, and there are several changes worth paying attention to if you run reporting or data tools in your business.
I’ve been working through the update this week and wanted to share what stood out.
The February release touches several parts of the platform, from how you filter and paste data into reports through to how Copilot reads and summarises your apps.
Below I’ll walk through the Input Slicer, Conversational Copilot, card visual updates, new DAX functions, and some important deprecation notices with deadlines coming mid-2026.
If you use Power BI day-to-day, there are a few things here worth acting on.
Ditch the Slicer Struggle
Picture this: you’re in a board meeting and the sales director emails a list of 50 priority clients from Queensland. Normally, you’d spend 20 minutes manually selecting them in Power BI slicers, one by one, while the room waits. That’s a frustrating and avoidable problem.
The new Input Slicer, now generally available (GA), addresses this directly. You type or paste free-text values like customer IDs, product codes, or postcodes, and it filters using “contains any” or “starts with” logic.
It also works as a task flow input for write-backs, which means it can feed data directly into Microsoft Fabric workflows. (Fabric is Microsoft’s unified data and analytics platform, bringing together data storage, reporting, and AI in one place.)
This means instant regional filtering without building custom visuals.
Paste into Slicers takes it one step further. Copy a list from Excel, an email, or a Teams chat and drop it straight into the slicer. No retyping required.
Copilot Gets Conversational
Fabric Copilot capacities are now enabled by default, tenant-wide, from 8 February 2026. No configuration required; your teams get AI-assisted reporting out of the box.
The headline feature here is Conversational Copilot in Power BI apps.
Rather than requiring you to pre-select a report, Copilot scans your entire app, finds the relevant data, and returns summaries or visuals based on plain-language questions like “Show QLD revenue trends vs last year.
If the data is ambiguous, it asks a clarifying question before responding, which keeps results accurate rather than just fast.
I’ve seen this work well for hybrid Australian teams. Executives on the Gold Coast can query “Melbourne client churn” from their phone and get a clear visual without waiting on a BI developer. The Power BI app summaries also now auto-generate overviews that can flag spending spikes or unusual patterns in your data.
For businesses I work with across manufacturing and services, this is also genuinely useful. Non-technical managers get the same quality of reporting that used to sit only with senior analysts, and it reduces pressure on already-stretched IT teams.
Cards That Command Attention
The Card visual has received a significant update. It now defaults to 10 callout items, covering KPIs like revenue, margins, and headcount, with interactive category filtering built in.
Click a category and it drills down, with no extra visuals needed.
Cross-platform rendering has also been improved, so your Brisbane dashboard displays correctly on a MacBook in a Melbourne meeting room.
Tooltips are cleaner too. Single-value cards show trend indicators with directional icons, while multi-row cards can display category images and top-N records (a filtered list of your highest-priority items).
As an example, a card showing “AU Sales: $2.4M (+12%)” with a green up-arrow could then filter to “QLD only” to reveal which regions are performing.
For our clients focused on Microsoft 365 integration, these card visuals pair well with Copilot. A card showing Fabric lakehouse metrics (your centralised data storage layer) can trigger a Copilot query like “Why did costs rise?” and bring the answer back into the same view.
Good News for DAX Developers
If your team builds custom data models, the new DAX functions NAMEOF and TABLEOF will save a lot of rework. DAX (Data Analysis Expressions) is the formula language used in Power BI to build calculations and data models.
Previously, if a column or table was renamed, any formulas referencing it would break. With NAMEOF and TABLEOF, formulas automatically adapt when names change. You write the reference once and it survives renames.
TABLEOF works the same way for full tables, which is particularly useful for user-defined functions (custom calculations you build and reuse) in Fabric semantic models. A semantic model is the organised data layer that Power BI reports read from.
This also benefits businesses connecting .NET or Python applications via the new ODBC driver preview.
Australian businesses running hybrid data environments, with both on-premises SQL databases and Azure cloud data, can avoid report refresh failures when their data schema changes.
We use this approach in Fabric migration projects to keep BI assets working reliably as businesses scale.
Sharing Reports Just Got Better
PBIR (Power BI Report format) is now the default export format. This means reports shared between team members retain full interactivity, rather than being flattened into PDFs that can’t be filtered or explored.
Azure Maps visuals also received a performance update this month, with faster rendering when working with pie chart overlays. Azure Maps now supports two new endpoint regions, Korea and Brazil, giving businesses more flexibility on where their data is processed.
Copilot Governance Controls
The Fabric Copilot capacity tenant setting is now enabled by default from 8 February 2026.
Administrators can manage which capacities are designated for Copilot use, giving IT teams control over how and where AI features are rolled out across the business.
Action Required: Legacy Excel and CSV Import Being Retired
The legacy Excel and CSV import experience in Power BI Service is being retired. Entry points will be removed on 31 May 2026, and semantic models created using that legacy import will stop refreshing on 31 July 2026.
If your team still uses this approach, consider a move to Fabric’s OneLake.
OneLake is Microsoft’s unified storage layer that connects all your data in one place without duplicating it.
What This Means for Your Business
These updates are about reducing the manual work that sits between your data and your decisions.
Input slicers and paste functions cut the time spent setting up reports, freeing teams to focus on what the data actually means. Conversational Copilot means business leaders can get answers directly from their data without always needing a developer in the loop.
The use cases I keep seeing across our client base follow a similar pattern:
For Australian businesses, there’s an added layer of confidence here. Fabric runs on local Australian data centres, which means data stays onshore, performance is consistent, and businesses in regulated sectors like energy, healthcare, and government can meet local data handling requirements without additional configuration.
Your Next Steps
If you want to start putting these features to work, here’s where I’d begin:
If you’d like to talk through how these updates apply to your business, feel free to reach out via LinkedIn or click the below to book a discovery session with one of our consultants.
About the Author
Carlos Garcia is the Founder and Managing Director of CG TECH, where he leads enterprise digital transformation projects across Australia.
With deep experience in business process automation, Microsoft 365, and AI-powered workplace solutions, Carlos has helped businesses in government, healthcare, and enterprise sectors streamline workflows and improve efficiency.
He holds Microsoft certifications in Power Platform and Azure and regularly shares practical guidance on Copilot readiness, data strategy, and AI adoption.
Sources
Recent Posts
Popular Categories
Archives