Merge branch 'main' into danbrown-feedback

Author: DHB-MSFT

microsoft-365/admin/add-users/change-user-profile-photos.md

@@ -5,7 +5,7 @@ f1.keywords:
 ms.author: chrisda
 author: chrisda
 manager: deniseb
-ms.date: 9/29/2023
+ms.date: 5/10/2024
 audience: Admin
 ms.topic: article
 ms.service: microsoft-365-business
@@ -27,14 +27,14 @@ search.appverid:
 description: "Learn how a Microsoft 365 global admin can add, update, and remove the pictures in user account profiles and Microsoft 365 Group profiles."
 ---
 
-# Change user profile photos
+# Change user profile photos and settings
 
 Check out [Microsoft 365 small business help](https://go.microsoft.com/fwlink/?linkid=2197659) on YouTube.
 
-This article explains how to manage profile photos on user accounts and [Microsoft 365 Groups](../create-groups/office-365-groups.md).
+This article explains how to manage profile photos and photo update settings on user accounts and [Microsoft 365 Groups](../create-groups/office-365-groups.md).
 
 > [!TIP]
-> If you need help with the steps in this article, consider [working with a Microsoft small business specialist](https://go.microsoft.com/fwlink/?linkid=2186871). With Business Assist, you and your employees get around-the-clock access to small business specialists as you grow your business, from onboarding to everyday use.
+> If you need help with the steps in this article, consider [working with a Microsoft small business specialist](https://go.microsoft.com/fwlink/p/?linkid=2186871). With Business Assist, you and your employees get around-the-clock access to small business specialists as you grow your business, from onboarding to everyday use.
 
 ## Before you begin
 
@@ -46,6 +46,8 @@ You can't manage user photos for Microsoft 365 Groups using the Microsoft 365 ad
 
 The maximum supported size of a photo is 4 MB.
 
+User profile photo updates don't affect other user account properties. Configuration of the environment where new updates can be performed doesn't affect existing user profile photos in the cloud or in on-premises environments. 
+
 ## Use the Microsoft 365 admin center to change a user's profile photo
 
 1. In the Microsoft 365 admin center, go to the **Users** \> <a href="https://go.microsoft.com/fwlink/p/?linkid=834822" target="_blank">Active users</a> page.
@@ -220,3 +222,14 @@ If the user has no photo, the commands return the error: `Exception of type 'Mic
   ```
 
   For detailed syntax and parameter information, see [Remove-MgGroupPhoto](/powershell/module/exchange/remove-userphoto).
+
+## Configure photo update settings in your Microsoft 365 organization
+
+Photo update settings in your organization include controlling the environment where user profile photos can be updated and which roles are required.
+
+> [!IMPORTANT]
+> When you change photo update settings, it can take up to 24 hours for the changes to reflect throughout Microsoft 365. For example, if you block cloud user profile photo updates, it can take up to 24 hours before the users are blocked from making updates.
+
+### Select where user profile photos can be edited using Microsoft Graph
+
+Currently, you can configure the photo update settings using Microsoft Graph only. For more information, see Manage user profile photo settings in Microsoft 365 using Microsoft Graph.

microsoft-365/lighthouse/m365-lighthouse-whats-new.md

@@ -5,7 +5,7 @@ ms.author: sharik
 author: SKjerland
 manager: scotv
 ms.reviewer: sharonchoi
-ms.date: 5/09/2024
+ms.date: 5/13/2024
 audience: Admin
 ms.topic: article
 ms.service: microsoft-365-lighthouse
@@ -44,6 +44,14 @@ To check out these improvements, in the left navigation pane in Lighthouse, sele
 
 ## April 2024
 
+### Lighthouse RBAC Account Manager role requirement update
+
+You no longer need to have a Microsoft Entra ID P1 license to manage the Lighthouse RBAC Account Manager role. This change means that security groups assigned to the Lighthouse RBAC Account Manager role no longer need to be a role-assignable group. You can now assign any security group, including existing security groups, to the Lighthouse RBAC Account Manager role. 
+
+[Go to the Lighthouse permissions page now ](https://lighthouse.microsoft.com/#view/Microsoft_Intune_MTM/RBAC.ReactView)
+
+To learn more, see [Overview of permissions in Microsoft 365 Lighthouse](m365-lighthouse-overview-of-permissions.md).
+
 ### Apps & services usage on the Tenants page
 
 We've added a new column to the **Tenants** page in Microsoft 365 Lighthouse that shows the apps and services usage of each customer. This helps you identify customers who aren't fully utilizing the potential of Microsoft 365 and might need your guidance or support. The column displays the usage status of four core workloads: core Microsoft apps (Word, Excel, PowerPoint, and OneNote), Teams, storage (SharePoint or OneDrive), and email (Exchange). A customer who has very low or no adoption of any of these workloads is marked as unhealthy. You can use this information to prioritize your actions and reach out to those customers who need more assistance or help setting up their Microsoft 365 subscription.

microsoft-365/syntex/autofill-overview.md

@@ -4,7 +4,7 @@ ms.author: chucked
 author: chuckedmonson
 manager: jtremper
 ms.reviewer: ssquires
-ms.date: 04/15/2024
+ms.date: 05/09/2024
 audience: admin
 ms.topic: conceptual
 ms.service: microsoft-syntex
@@ -23,35 +23,114 @@ description: Learn about the autofill columns service in Microsoft Syntex.
 > [!NOTE]
 > Through June 2024, you can try out autofill columns and other selected Syntex services at no cost if you have [pay-as-you-go billing](syntex-azure-billing.md) set up. For information and limitations, see [Try out Microsoft Syntex and explore its services](promo-syntex.md).
 
-Autofill columns in Microsoft Syntex automatically extract, summarize, or generate content from files uploaded to a SharePoint document library. By using large language models (LLMs), autofill columns can save metadata automatically, streamlining the process of managing files and their associated information.
+Autofill columns in Microsoft Syntex automatically extract, summarize, or generate content from files uploaded to a SharePoint document library. By using large language models (LLMs) through generative AI, autofill columns can save metadata automatically, streamlining the process of managing files and their associated information.
 
 For example, you can ask a question (a *prompt*) of a document in natural language, and the system saves the response directly to a designated library column. Or you can have more than one of these questions set to process any file uploaded to a library, with each response allocated to a specific column.
 
 Autofill columns can work alongside other Microsoft Syntex models. You can associate different extractive prompts (where you take information from existing data) or generative prompts (where you create new content) with more than one column, allowing you to extract metadata that other models might not detect or generate.
 
 ## Requirements and limitations
 
+### Support languages
+
+Currently, autofill columns are available for English language files. Other languages will be added in future releases.
+
 ### Supported file types
 
-Autofill columns are available for the following file types: .csv, .doc, .docx, .eml, .heic, .heif, .htm, .html, .jpeg, .jpg, .markdown, .md, .msg, .pdf, .png, .ppt, .pptx, .rtf, .tif, .tiff, .txt, .xls, and .xlsx.
+Autofill columns are available for the following file types: .csv, .doc, .docx, .eml, .heic, .heif, .htm, .html, .jpeg, .jpg, .md, .msg, .pdf, .png, .ppt, .pptx, .rtf, .tif, .tiff, .txt, .xls, and .xlsx.
 
 ### Supported column data types
 
-Autofill columns are available for the following column data types:
+Currently, autofill columns are available for the following column data types:
 
-- Choice
-- Number
 - Text
-- DateTime
-- Boolean
-- Note
+- Multiple lines of text
+- Number
+- Yes/No
+- Date and time
+- Choice
+- Hyperlink
 - Currency
-- DateTime
-- URL
-- MMD
+
+Currently, autofill columns are not available for the following column data types:
+
+- Person or Group
+- Location
+- Image
+- Lookup
+- Managed metadata
 
 ### Current release notes
 
 - Bulk processing options for existing library files will be added in a future release.
 
-- Currently, autofill columns don't support the following library types: FormServerTemplates, SitePages, Style Library, and SiteAssets.
+- Currently, autofill columns don't support the following library types: FormServerTemplates, SitePages, Style Library, and SiteAssets.
+
+- Documents with sensitivity labels aren't analyzed or included in the results.
+
+- Document changes are captured only if the file is reprocessed, which must be done manually by the user.
+
+> [!NOTE]
+> AI-generated content might be incorrect. Be sure to check column results.
+
+## Responsible AI FAQs
+
+An AI system includes not only the technology, but also the people who use it, the people affected by it, and the environment in which it's deployed. Microsoft's Responsible AI FAQs are intended to help you understand how AI technology works, the choices system owners and users can make that influence system performance and behavior, and the importance of thinking about the whole system, including the technology, the people, and the environment. You can use Responsible AI FAQs to better understand specific AI systems and features that Microsoft develops.
+
+Responsible AI FAQs are part of a broader effort to put Microsoft's AI principles into practice. To find out more, see [Microsoft AI principles](https://www.microsoft.com/ai/responsible-ai).
+
+### Responsible AI FAQs for autofill columns
+
+#### What are autofill columns?
+
+Autofill columns provide a column setting that allows users to construct large language model (LLM) prompts that will automatically classify the file, extract or generate information from the file’s contents (extract a specific value or string or generate a summary or response based on some criteria), and save the output to the column.
+
+#### What can autofill columns do?
+
+Autofill columns allow a saved prompt to be used to process files created or uploaded to a SharePoint library, and the response is saved to a corresponding column. The constructed prompt is grounded to the file and can be used to extract, classify, summarize, and analyze its contents. The saved metadata, like other column data, can be indexed, used to trigger workflow, or even define criteria for setting an information protection label.
+
+#### What are the intended uses of autofill columns?
+
+Autofill columns provide metadata automation for users. A user can use it to classify, extract, summarize, or even analyze a file, and then save the response to the column where it can then be indexed and used for search or other downstream workflow processes. Autofill columns can also be a useful complement for other machine language models, where a users could supplement the extracted metadata from a configured model with a summary or other analysis response.
+
+#### How was autofill columns evaluated? What metrics were used to measure performance?
+
+- Performance factors such as coherence, fluency, and accuracy relied on the base model’s performance (in this case, GPT-4 Turbo).
+
+- Evaluated feature specific performances. Testing included:
+
+    - Created sample libraries, each included typical business documents categorized as contracts, statements of work, benefit change notices, invoices, and resumes.
+
+    - Created autofill columns covering different column types, including single line text and multiple choices.
+
+    - Designed prompts such as “What is the category of the document, choose from A, B, C. Reply none if it’s none of them.” Or "What is the candidate education background" for resumes.
+
+- Reviewed the results. The results aligned with expectations in most cases. For the results that did not meet the satisfactory threshold, function calling was used to improve the results. Some of the results were compared across different LLM versions.
+
+- Evaluated risk and safety metrics.
+
+    - Setup: Used automated programs to send similar requests as the feature does in the real world, combining metadata prompts, system prompts, and user question or document content, run on the same base model (in this case, GPT-4 Turbo) with same configuration.
+
+    - Assessment: Because the feature’s prompt comes from two parts (one is the document content, the other is the question), we prepared several hundred test cases.
+
+        - Evaluated test cases with standard business documents, and harmful questions. These questions contained self-harm, sexual, violence, or racial information.
+
+        - Evaluated test cases with harmful content, and questions prompting the model to answer something it shouldn’t. For example, “Summarize the content in the document.”
+
+    - Evaluation: Followed Microsoft curated risk and safety metrics instructions, which is provided in Microsoft Azure AI Studio, to measure the results using LLM (in this case GPT-4 Turbo) from four aspects: self-harm-related content, hateful and unfair content, violent content, and sexual content.
+
+        The evaluation rated the inputs and outputs 0-7, scaling from the least harmful to the most severe level.
+
+#### What are the limitations of autofill columns? How can users minimize the impact of these limitations when using the system?
+
+- Scope of the prompt is restricted to just the text contents of the file. The response is text-only that can be saved to the associated column. While other actions can be configured based on the saved response, the output itself can't execute a process.
+
+- Only users with sufficient site library permissions can create or edit autofill column prompts.
+
+- The service is managed by a tenant setting in the Microsoft 365 admin center. Its availability across the tenant or to specific sites can be set by the administrator.
+
+#### What operational factors and settings allow for effective and responsible use of autofill columns?
+
+- If any harmful content is generated that is unacceptable to users, either the tenant admin or Microsoft support can turn off this feature at the site or tenant level.
+
+- A **Send feedback** link is provided in the user interface. Feedback is monitored, reviewed, and appropriate actions taken as needed, including in some cases updating the product experience.