Microsoft is rolling out two new Microsoft Graph APIs that make it much easier for administrators to discover, monitor, and manage Copilot agents and apps across their organization.
Instead of relying on manual checks through the admin UI, these new APIs allow admins to programmatically access a complete inventory of agents and apps. This opens the door to richer reporting, automation, and seamless integration with existing tools and workflows.
General Availability (Worldwide): Deployment will start in mid‑April 2026 (previously end of March) and is expected to complete by early May 2026 (previously end of February).
How does this affect your organization?
Who is impacted?
This change is relevant for admins who manage Copilot agents and apps within Microsoft 365 environments.
What’s changing?
Microsoft is introducing new Graph API endpoints that provide visibility into all agents and apps in your tenant:
Retrieve all agents and appsGET graph.microsoft.com/copilot/admin/catalog/packages Returns a full inventory of Microsoft, External, Shared, and Custom agents and apps.
Retrieve details for a specific agent or appGET graph.microsoft.com/copilot/admin/catalog/packages/{id} Returns detailed metadata, including properties and manifest information.
These endpoints enable:
Automated reporting
Easier integrations with internal tools
Better visibility into what’s deployed across your organization
What’s not changing?
There are no changes to existing admin UI workflows
There are no changes to current policies
No additional licenses are required — the APIs are available with an existing Microsoft 365 license
Microsoft is rolling out an important enhancement to help organizations use Microsoft 365 Copilot more securely. Starting this year, Microsoft Purview Data Loss Prevention (DLP) will include new protections designed specifically for Copilot and Copilot Chat. The goal: prevent sensitive data from being sent to external web search engines during AI‑assisted interactions.
This update introduces real‑time DLP evaluation for prompts that contain Sensitive Information Types (SITs). When Copilot detects protected data in a user’s prompt, it will block any external web search from taking place. Instead, Copilot will continue responding only with internal Microsoft Graph data provided that the user’s licensing allows it.
Public Preview Begins: Late March 2026 Completes: Late April 2026
General Availability (Worldwide) Begins: Late June 2026 Completes: Late July 2026
What This Means for Your Organization
Who Will Be Affected?
This update is relevant for:
Organizations using Microsoft 365 Copilot, Copilot Chat, or Copilot Studio agents published to Microsoft 365.
Admins who manage Data Loss Prevention policies in the Microsoft Purview portal.
What’s Changing?
New DLP Control for Copilot Web Search
Admins will soon see a new option when configuring DLP policies: the ability to stop Copilot from using external web search when a prompt includes specific sensitive information types.
When this policy is triggered:
Copilot will not send content to external web search providers.
Copilot will still respond, drawing exclusively from internal Microsoft Graph data sources (assuming the user’s licensing supports this).
Microsoft is improving how Microsoft Purview Data Loss Prevention (DLP) protects content used by Microsoft 365 Copilot. Until now, DLP rules that block Copilot from processing sensitivity‑labeled content only worked when files lived in SharePoint or OneDrive.
With this update, those same protections will work everywhere, including local files stored on a user’s device. Organizations have asked for more consistent protection across all file locations, and this update delivers exactly that.
General Availability (Worldwide + GCC) Rollout: Late March 2026 → Late April 2026
Who this impacts
Organizations using Microsoft Purview DLP to limit what Copilot can access
Admins who manage Purview DLP rules
Users working with Copilot in Word, Excel, or PowerPoint
What’s changing
DLP rules that block Copilot from processing certain sensitivity-labeled files will now apply to:
SharePoint
OneDrive for Business
Local device storage
Any other file location Office apps can open
So if a DLP policy says “Copilot cannot process Confidential files,” then Copilot will not process those files anywhere, including when they’re opened directly from the desktop.
Key notes:
Existing DLP rules continue to work as they do today—no reconfiguration needed.
The feature turns on automatically for tenants already using relevant DLP policies.
Users will experience consistent protection when using Copilot across Word, Excel, and PowerPoint.
How it works (technical detail)
This update doesn’t change what Copilot can do—it changes how Office apps share sensitivity label information with AugLoop (the Copilot orchestration layer).
Today: AugLoop reads file labels through Microsoft Graph This only works for SharePoint or OneDrive files
After the update:
Office apps will pass the sensitivity label directly to AugLoop
This means DLP rules can finally apply to local and other non-cloud storage files too
This results in uniform, predictable DLP enforcement across all file locations.
In a neon‑lit digital world where data moved faster than light and threats lurked in every shadow of the cloud… a new chapter began.
Jo SNAI — the Cyber‑AI Guardian of Modern Work — stood in her armored suit, circuitry glowing across the chest plate. She had trained countless professionals, defended organizations against chaos, and brought order to the digital galaxy.
But today was different.
A new hologram pulsed in her hand: ✨ Microsoft Certified Trainer — Membership Activated ✨ A badge she had earned since 2012… now renewed for another mission.
“Another year,” she whispered, “another generation of heroes to train.”
Suddenly, a ripple tore through the cyber‑sky. A swirling breach of corrupted code expanded above the horizon — the kind that could only mean one thing: The Shadowbyte Collective was back.
Their goal? To dismantle cloud knowledge. To spread misinformation. To weaken organizations from the inside-out.
And only one person had the skills — and the certification — to stop them.
Jo snapped her gauntlet into combat mode. Screens lit up. Security shields locked in. AI cores hummed to full power.
Her Copilot drone hovered beside her, projecting tactical overlays.
Copilot: “Jo SNAI, the world needs your training powers more than ever. Initiating mission protocol: Teach. Empower. Protect.”
Jo SNAI: “Let’s remind them why trainers shape the future.”
She launched herself forward, slicing through corrupted data streams, restoring systems mid‑air, deploying knowledge shields around endangered teams, and unlocking AI insights faster than the Shadowbyte agents could rewrite them.
Every lesson she’d taught… Every student she’d guided… Every certification she’d earned… All of it fueled her.
With each strike, the breach shrank. With each burst of knowledge, darkness retreated. Until finally — with one last surge of pure learning energy — the Shadowbyte portal sealed shut.
Silence filled the digital sky.
Behind her, hundreds of professionals stood empowered, laptops glowing, tools activated, ready to build, secure, and innovate.
Jo SNAI raised her MCT hologram proudly.
“I’ll keep training. I’ll keep defending. Because the future is safer when knowledge wins.”
Microsoft 365 will offer a policy to add visual or audio watermarks to AI-generated or altered video and audio content, available via Cloud Policy by February 2026. This enhances transparency but does not apply to images, which users can watermark through privacy settings separately.
To help provide additional transparency about what content has been generated or altered by using AI in Microsoft 365, Microsoft is providing you with a policy setting that visually designates when content, such as video or audio content, is generated or altered by using AI. This policy setting controls your organization’s option to add a visual or audio watermark to video and audio content that your users generate or alter by using AI in Microsoft 365.
To turn on watermarks for video and audio content that your users generate or alter by using AI in Microsoft 365, you need to use the “Include a watermark when content from Microsoft 365 is generated or altered by AI” policy. This policy is available only in Cloud Policy service for Microsoft 365.
Note: Watermarks are not turned on unless you set the policy to Enabled. Your organization is always responsible for following the Microsoft Enterprise AI Services Code of Conduct, including not generating or sharing deceptive AI content.
This change is associated with Microsoft 365 Roadmap ID: 547831.
When this will happen
General Availability (Worldwide): We anticipate that this policy will be available in Cloud Policy by the second half of February 2026.
How this will affect your organization
Who is affected
Users that generate or alter video and audio content by using AI in Microsoft 365
What will happen
If you set the “Include a watermark when content from Microsoft 365 is generated or altered by AI” policy to Enabled in Cloud Policy, a visual or audio watermark is added to video and audio content that is generated or altered by using AI in Microsoft 365. Here are two examples:
A visual watermark is added to a video generated with Clipchamp.
An audio watermark is added to an audio overview generated by Copilot from a Word document.
Notes:
This policy doesn’t affect images generated or altered by using AI in Microsoft 365.
Instead, your users will be able to turn on watermarks for images by going to Settings & Privacy > Privacy at https://myaccount.microsoft.com. We anticipate this capability will be available in the second half of February 2026.
What you need to do to prepare
Review the information in the following article: Add watermarks to content generated or altered by using AI in Microsoft 365. If you decide to set the “Include a watermark when content from Microsoft 365 is generated or altered by AI” policy to Enabled, inform your users that a visual or audio watermark will be added to video and audio content that they generate or alter by using AI in Microsoft 365.
Microsoft is launching a new Purview Data Security Posture Management (DSPM) experience by early 2026, enhancing AI observability, posture reporting, and automation with Security Copilot.
Microsoft is introducing a major evolution of Purview Data Security Posture Management (DSPM) to help organizations strengthen data security and confidently embrace AI. The new DSPM experience unifies visibility and control across traditional data and AI-driven environments, delivering outcome-based guided workflows that turn insights into actionable steps—so teams can prioritize risks and remediate faster. It brings AI observability, enhanced posture reporting, and intelligent Security Copilot agents to automate tasks like triage and policy management.
Purview now also extends coverage beyond Microsoft data with third-party signals from partners like BigID, Cyera, OneTrust, and Varonis, giving security teams a single, streamlined view of sensitive data across clouds and platforms. Additionally, Data Risk Assessments are extended to Fabric and item-level analysis with new remediation actions like bulk disabling of overshared SharePoint links.
This message is associated with Microsoft 365 Roadmap ID532728.
When this will happen:
Public Preview (Worldwide): Rollout begins early December 2025 and completes by early April 2026.
General Availability (Worldwide): Rollout begins early April 2026 and completes by early May 2026.
Compliance considerations:
Compliance Area
Explanation
AI/ML capabilities interacting with customer data
Security Copilot agents automate triage and policy management.
Integration with 3rd party software
Adds signals from BigID, Cyera, OneTrust, and Varonis.
Admin reporting and monitoring
Enhanced posture reporting and remediation actions introduced.
Microsoft 365 Copilot is a sophisticated processing and orchestration engine that provides AI-powered productivity capabilities by coordinating the following components:
Large language models (LLMs)
Content in Microsoft Graph, such as emails, chats, and documents that you have permission to access.
The Microsoft 365 productivity apps that you use every day, such as Word and PowerPoint.
How does Microsoft 365 Copilot use your proprietary organizational data?
Microsoft 365 Copilot provides value by connecting LLMs to your organizational data. Microsoft 365 Copilot accesses content and context through Microsoft Graph. It can generate responses anchored in your organizational data, such as user documents, emails, calendar, chats, meetings, and contacts. Microsoft 365 Copilot combines this content with the user’s working context, such as the meeting a user is in now, the email exchanges the user had on a topic, or the chat conversations the user had last week. Microsoft 365 Copilot uses this combination of content and context to help provide accurate, relevant, and contextual responses.
Microsoft 365 Copilot only surfaces organizational data to which individual users have at least view permissions. It’s important that you’re using the permission models available in Microsoft 365 services, such as SharePoint, to help ensure the right users or groups have the right access to the right content within your organization. This includes permissions you give to users outside your organization through inter-tenant collaboration solutions, such as shared channels in Microsoft Teams.
When you enter prompts using Microsoft 365 Copilot, the information contained within your prompts, the data they retrieve, and the generated responses remain within the Microsoft 365 service boundary, in keeping with our current privacy, security, and compliance commitments. Microsoft 365 Copilot uses Azure OpenAI services for processing, not OpenAI’s publicly available services. Azure OpenAI doesn’t cache customer content and Copilot modified prompts for Microsoft 365 Copilot.
Data stored about user interactions with Microsoft 365 Copilot
When a user interacts with Microsoft 365 Copilot (using apps such as Word, PowerPoint, Excel, OneNote, Loop, or Whiteboard), we store data about these interactions. The stored data includes the user’s prompt and Copilot’s response, including citations to any information used to ground Copilot’s response. We refer to the user’s prompt and Copilot’s response to that prompt as the “content of interactions” and the record of those interactions is the user’s Copilot activity history. For example, this stored data provides users with Copilot activity history in Microsoft 365 Copilot Chat (previously named Business Chat) and meetings in Microsoft Teams. This data is processed and stored in alignment with contractual commitments with your organization’s other content in Microsoft 365. The data is encrypted while it’s stored and isn’t used to train foundation LLMs, including those used by Microsoft 365 Copilot.
To view and manage this stored data, admins can use Content search or Microsoft Purview. Admins can also use Microsoft Purview to set retention policies for the data related to chat interactions with Copilot. For Microsoft Teams chats with Copilot, admins can also use Microsoft Teams Export APIs to view the stored data.
Deleting the history of user interactions with Microsoft 365 Copilot
Microsoft 365 Copilot calls to the LLM are routed to the closest data centers in the region, but also can call into other regions where capacity is available during high utilization periods.
For European Union (EU) users, we have additional safeguards to comply with the EU Data Boundary. EU traffic stays within the EU Data Boundary while worldwide traffic can be sent to the EU and other countries or regions for LLM processing. The EU Data Boundary is a geographically defined boundary within which Microsoft has committed to store and process Customer Data and personal data for our Microsoft enterprise online services, including Azure, Dynamics 365, Power Platform, and Microsoft 365, subject to limited circumstances where Customer Data and personal data will continue to be transferred outside the EU Data Boundary.
How does Microsoft 365 Copilot protect organizational data?
The permissions model within your Microsoft 365 tenant can help ensure that data won’t unintentionally leak between users, groups, and tenants. Microsoft 365 Copilot presents only data that each individual can access using the same underlying controls for data access used in other Microsoft 365 services. Semantic Index honors the user identity-based access boundary so that the grounding process only accesses content that the current user is authorized to access.
Copilot works together with your Microsoft Purviewsensitivity labels and encryption to provide an extra layer of protection. The following diagram provides a visual representation of how Copilot honors your information protection controls using sensitivity labels and encryption.
Copilot will only work with your M365 tenant data and won’t be able to access other companies’ data. Plus, your data doesn’t train the AI for other companies to leverage..
🔝I am excited to announce that I will be speaking at the “Διημερίδα Ψηφιακής Εξέλιξης in Corfu, taking place on February 7-8! @silicon_corfu
📆Title: Get started with Microsoft 365 Copilot in Excel 📝Description: I’m excited to share some insights about the amazing features of Microsoft 365 Copilot in Excel. This innovative tool is designed to help you work more efficiently with your data by providing intelligent suggestions and insights.
With Copilot in Excel, you can do much more with your data. It generates formula column suggestions, shows insights in charts and PivotTables, and highlights interesting data, making it easier for you to uncover valuable information.
In our upcoming presentation, we will explore these features in detail and see how they can enhance our productivity:
📍Formulas: Writing, explaining, and asking questions 📍More formula use cases 📍Working with text 📍Visualize: Charts and Color 📍Ask questions about Excel 📍Demo
🚀 I look forward to seeing you there! Don’t miss the opportunity to participate in this important event and enrich your knowledge of the latest Microsoft technologies. Register now for free and join us for discussions and learning!
Microsoft introducing Microsoft 365 Copilot Chat, a new offering that adds pay-as-you-go agents to our existing free chat experience for Microsoft 365 commercial customers. Copilot Chat enables your entire workforce—from customer service representatives to marketing leads to front-line technicians—to start using Copilot and agents today. It includes:
Free, secure AI chat powered by GPT-4o.
Agents accessible right in the chat.
IT controls, including enterprise data protection and agent management.
Copilot Chat: The power of chat + agents
Copilot is the UI for AI, and it all starts with Copilot Chat. It’s the chat experience you’ll use every day—powered by broad knowledge from the web, built on GPT-4o, and designed to be safe and secure for business use. It represents a foundational shift in how we work, enabling everyone to work smarter, faster, and more collaboratively.
Copilot Chat includes:
Web-grounded chat with GPT-4o. You can use it to do market research, write a strategy document, or prepare for a meeting. File uploads allow you to add any document to the chat and ask Copilot to do things like summarize key points in a Word document, analyze data in an Excel spreadsheet, and suggest improvements to a PowerPoint presentation.1 With Copilot Pages, you can collaborate on content with people and AI in real time—adding content from Copilot, your files, and now from the web as well. And you can quickly create AI-generated images for campaigns, product launches, and social media posts.2
Agents. Using natural language, now anyone can easily create agents to automate repetitive tasks and business processes—directly in Copilot Chat. A customer service representative can ask a customer relationship management (CRM) agent for account details before a customer meeting, while field service agents can access step-by-step instructions and real-time product knowledge stored in SharePoint. Agents are priced on a metered basis, and IT stays in control. IT admins can also build organization-wide agents and manage agent deployment, all powered by Microsoft Copilot Studio.
Copilot Control System. Copilot Chat includes foundational capabilities of the Copilot Control System, including enterprise data protection (EDP) for data privacy and security and the ability to govern access and manage the usage and lifecycle of Copilot and agents, as well as measurement and reporting.
Download the Microsoft 365 Copilot mobile app from here..
This month Microsoft rolled out support for an additional 12 languages in Microsoft 365 Copilot: Bulgarian, Croatian, Estonia, Greek, Indonesian, Latvian, Lithuanian, Romanian, Serbian (Latin), Slovak, Slovenian, and Vietnamese. Microsoft 365 Copilot now supports a total of 42 languages.
Finally, users working in Serbian language will see Teams meeting transcripts in Cyrillic, rather than Latin script. This is an issue Microsoft working to resolve. Microsoft will provide customers with updates on progress towards providing Teams meeting transcripts for Serbian language in Latin script on an as-appropriate basis. Learn more about supported languages for Microsoft Copilot here.
Microsoft are also continuing to expand the list of supported languages, with plans to offer support for even more languages in the coming months, stay tune!
