Strengthen your security posture with Microsoft Entra Conditional Access
Learn how Microsoft Entra Conditional Access, our Microsoft Zero Trust policy engine, protects access for your workforce and for agents by enforcing real‑time adaptive access policies that continuously assess risk signals and use AI‑driven automation to dynamically allow, challenge, or block access for every identity. Join Microsoft experts as they walk through real‑world scenarios and share practical guidance to help your identity team address policy sprawl, enforce consistent Conditional Access policies, and strengthen security posture across your environment. How do I participate? Registration is not required. Add this event to your calendar, then sign in to the Tech Community and select Attend to receive reminders. Post your questions in advance, or any time during the live broadcast. Note: This session was originally scheduled for June 8, 2026 and will now take place on June 24, 2026.
Why “Data in Switzerland” Is Not Enough
Moving from Residency to Control in Microsoft 365 Every conversation about data sovereignty in regulated industries tends to start the same way: “We use Multi-Geo. The data stays in Switzerland.” It’s the right starting point. Microsoft 365 Multi-Geo allows organizations to place selected workloads - SharePoint sites, OneDrive accounts, Teams data, or Exchange mailboxes - into specific regions, including Switzerland, while maintaining a single global tenant. This makes it possible to align sensitive data with regulatory or customer requirements without fragmenting the overall environment. But it only answers one question: Where is the data stored? It does not answer who accessed the data, from where, under which conditions, or what happened after access. That is where the real problem begins. A scenario that happens every day A Swiss engineering firm stores sensitive project documentation in Switzerland using Multi-Geo. An external contractor - working from an unmanaged device outside Switzerland - is granted access to review a file. The document opens. The data is now on a screen in an unknown location, on a device with no compliance posture, in a session with no restrictions. From the platform’s perspective, residency was enforced. From a sovereignty perspective, control was lost the moment access was granted without conditions. The file never left Switzerland. But sovereignty did. Residency is static. Control is not. The moment a document is opened, storage location stops being the relevant boundary. The file is no longer just “in Switzerland.” It moves instantly across endpoints and browsers, collaboration tools like Teams, external users and partners, and increasingly AI-driven contexts. The infrastructure remains unchanged. The data does not. From the platform’s perspective, everything is working as designed - access was granted, residency was enforced - and control was lost. Most “data in Switzerland” strategies fail at exactly this moment: when the data is used. The shift: from location to conditions If data sovereignty is the goal, the question must change. Not “Where is the data stored?” but: Under which conditions can data be accessed and used? This shift fundamentally changes the architecture. Control must be applied across three distinct layers - and all three must be connected. Layer 1: Access is conditional, not static Conditional Access extends control beyond authentication and turns it into continuous evaluation. Access decisions can depend on: Device compliance Location (geo-restriction) Identity and risk signals Multi-Geo ensures data is placed correctly. Conditional Access ensures it is reachable only under defined conditions. The two must work together - residency without access governance is an incomplete control. Layer 2: The session is the real risk surface Even with strict access controls, risk remains. A session is an exposure surface by design. During an active session, data is viewed, copied, shared, processed by applications, and connected to AI prompts. The gap does not appear at storage or authentication. It appears during active usage - inside the session. This is the layer most architectures do not explicitly address. Controls must extend into the session itself: limiting data transfer and replication, restricting interaction patterns, and enforcing policies in real time. Access is no longer a one-time event. It becomes continuously governed. This becomes even more critical as AI assistants consume content across SharePoint, Teams, Exchange, and other Microsoft 365 services. The question is no longer only where the source document resides - but whether the AI interaction itself is governed by the same access and protection controls as direct access. Layer 3: The document becomes the control point The most durable control does not sit in the network or in the session. It sits in the data itself. In regulated industries, organizations often arrive at this architecture having first evaluated sovereign or national encryption solutions. The decision to rely on native Microsoft 365 Purview encryption rather than a separate layer comes down to integration: AES-256 protection operating natively at file, user, and SharePoint level - including geo-based access restrictions - without an additional system to maintain. When protection is applied directly to the document through Microsoft Purview: Sensitivity labels define classification - automatically assigned based on content Encryption enforces access - AES-256, bound to the file itself IRM controls usage - view, copy, print, share, and presentation rights DLP governs movement across services - preventing data from leaving defined boundaries Dynamic watermarking tracks exposure - applied on open, view, or print At that point, access is enforced by the file, usage restrictions travel with it, and control persists regardless of location. The document becomes the perimeter. Platform control: limiting provider access One dimension often overlooked in sovereignty discussions is platform access itself. Even a perfectly configured tenant is only as sovereign as the controls placed on the operator. Customer Lockbox ensures that even Microsoft support cannot access customer data without explicit, logged, time-bound approval. Every access request is visible, auditable, and subject to customer veto. Data control applies not only to users - but also to the platform operating the service. Enforcement requires an integrated architecture Most organizations already have the required capabilities: Multi-Geo, Conditional Access, session control, Purview (labels, encryption, DLP, IRM), and monitoring. The issue is not capability. It is fragmentation. In practice, fragmentation looks like this: residency is configured in one project, Conditional Access policies are managed by a different team, and Purview labels were applied during a compliance initiative that never connected to the access layer. The tools exist. The signals do not flow between them. When designed as a single architecture: Data is placed intentionally - residency aligned to regulatory requirements Access is governed by context - device, location, and identity evaluated continuously Usage is controlled dynamically - session-level restrictions enforced in real time Protection is embedded in the document - encryption and IRM travel with the file Signals are connected across the platform - monitoring feeds access policy, not just audit logs “Data in Switzerland” becomes not just a statement - but an enforceable system property. Closing thought Placing data in Switzerland is the right first step. Multi-Geo makes it possible, even in global environments. But residency alone is not control. Data residency answers where information is stored. Data sovereignty requires proving who can access it, under which conditions, and what controls remain in place after access is granted. In Microsoft 365, sovereignty is no longer defined by geography alone. It is defined by the ability to enforce control wherever the data travels.Passkey Sign in Method (Entra Account) missing in Security
Hi Microsoft Support we enable FIDO2 passkey in entraId. However, when we try to register the FIDO2 passkey on myaccount.microsoft.com -> Security -> Add a Sign-in Method -> Passkey is missing. Attached screenshot. For a personal account, the Passkey method is available at the same location, even though interface is slightly different than an Entra Id account. Attached screenshot for the personal account as well. Kindly guide us on where to register the passkey or if we need to enable certain settings in EntraId for the passkey to show up in sign-in methods. We have Auth Strengths enabled in EntraId for the particular user in question and this reflects in the Device Lockscreen during login on Entra Registred Device. Thanks Chandra
Fido passkeys blocked by policy
Hi all I'm helping out a customer with deploying physical passkeys and I'm running into a weird error. I've activated the sign in method and selected the two AAGuids for the Authenticator app and I've added the right AAGuid for the brand and model of passkey we are using. We can select the authentication method and enroll the security correctly but when trying to sign in using it we get the error as displayed in the attached picture. When checking the sign in logs i get this error message FIDO sign-in is disabled via policy and the error code is: 135016 I've not been able to track down any policy that would be blocking passkeys. anyone got any ideas?App Enforced Restrictions not working on Chrome
Hi All I hope you are well. Anyway, a strange one here. We have implemented App Enforced Restrictions on unmanaged / BYOD macOS devices. This seems to have taken effect on Edge and Safari browsers but not Chrome. Is there anything we can do to resolve this or force BYOD macOS to use Edge? Info appreciated. SK
Policy applied allthough it shouldn't
Hi, all of a sudden Intune chaanges its behavior. I have a policy in place that sets persistent browser session. On the device filter tab I excluded devices with this syntax: device.trustType -eq "ServerAD" -or device.deviceOwnership -eq "Company" Starting last week I have to re-authenticate on a remote Desktop running Windows Server 2025 every 8 hours. Thats what the policy requires. In Entra I see in the logs for my user that this conditional access policy applied. I then extended the filter to this device.trustType -eq "ServerAD" -or device.deviceOwnership -eq "Company" -or device.operatingSystem -contains "Server" But it did not make a difference. Any idea what is going? This is not specific to my tenant. On a different tenant it behaves the same way.BYOD devices can't launch Windows 365 PC because of device compliance check during CA policy check.
We have a device compliance policy for all cloud apps. We would like to allow personal (BYOD) devices to be able to connect to Windows 365 Cloud PC. In the sign in logs we see the failures for application "Windows 365 Client" app id 4fb5cc57-dbbc-4cdc-9595-748adff5f414. We can't exclude that application in the conditional access policy as it's not available. We already added exclusions for Azure Virtual Desktop, Windows 365 and Windows Cloud Login. How can we allow BYOD devices to connect to cloud PCs?iOS managed contacts - how to deal with that?
Hi everyone, the last years i've already tried to solve the problem with the managed contacts. Because this was not possible earlier i forgot about that. Now i want to readress this issue. A very important article i've found is this one: Techcommunity Success: New contact sync scenario available with Outlook for iOS on enrolled devices With this thread i would like to discuss some unanswered questions of myself. I would really appreaciate any answer of you guys. 🙂 Goals: Business contacts should be able to be read through contacts app (because of caller-id) 3rd Party Messengers should not see these business contacts Thesises: It is not possible to achive this with Outlook for iOS and it's contact sync feature, right? (Because of these contacts are going to be synced through icloud, therefore these contacts are marked as "unmanaged contacts.) It is possible to achive these goals by using: an device configuration profile which configures an active sync account which only synchronizes the contacts of the users mailbox. These contacts are considdered as "managed contacts" an app configuration profile which disables the "sync contacts" feature for "outlook for ios" An App protection policy which disables "Viewing corporate documents in unmanaged apps Because of the fact this is only working for enrolled and managed devices, we need to tell the users: Caller identification is only possible if you enroll your device in Intune. (in relation to the previous points) So far, so good, but the bad news is: Because of the incopatibility with conditional access policies, we're hence not able to restrict the user from using other apps to connect their EXO account. Right? I would be very thankful if anyone can discuss this with me. (I think the best way to adress the different topics is to quote my post and answer inline.) Greetings, Patrick
