Adaptive Thinking - Frontier ReleaseAdaptive Thinking brings faster responses to lightweight interactions such as “Hi,” “What’s your name?” and “Thanks.” Instead of waiting for the full reasoning model, these simple turns can take a lightweight response path without unnecessary tool calls. More complex requests continue to use the full reasoning and tool-use pipeline.Rollout Schedule for next tiersStandard: November 5, 2026 Basic: November 10, 2026Technical RestrictionsNone.DocumentationNo help-site documentation is currently available.

The Moveworks Assistant in Slack now shows its work, formats answers more clearly, and allows administrators greater control over aspects like threading, Slack Work Objects, and native modal control. Most changes are on for everyone today. A few are optional settings your admins can turn on in Moveworks Setup.What has been enabled for all Moveworks instancesReasoning steps: while the Assistant works on a request, employees see a short, collapsible plan of what it's doing. Tables: tabular answers display as native Slack tables. Cleaner long answers: long answers display in full, without a Show more link. Channel Resolver messages: the message the Assistant sends when it responds to a channel post is written for that post, rather than templated Brief Me in threads: Brief Me stays focused on the thread you're in.You don't need to do anything to get these changes.Optional settings for adminsIn Moveworks Setup, under Chat Platforms > Manage Chatbots > your Slack bot:Allow Multiple Threads (Chats): employees can run more than one conversation at a time, each in its own thread. Enable Native Slack Modals: references, feedback and Get Help open as native Slack pop-ups. This allows users to  Enable Work Objects: ticket previews display in Slack. Update your Slack app before you turn this on.Under Channel Settings, edit the channel:Respond in Thread: Channel Resolver replies in the thread of the original post. Plugin Allowlist: choose which plugins Channel Resolver can use in that channel.Under Ticketing Automation > Handoff > Message Brokering:Enable Live Agent Handoff Summary: the live agent gets a summary of the conversation at handoff. Enable Post Session Feedback Card: employees can rate the live agent chat when it ends. Enable Agent Impersonation: live agent messages show the agent's name and photo. Add the chat:write.customize scope to your Slack app and reinstall it before you turn this on.Setup steps are in Configure Slack Features and Configure Channel Resolver.Good to knowAvailable in every Moveworks data center at no extra cost. Microsoft Teams is unchanged. Turn on Enable Agent Impersonation only after the Slack app has the chat:write.customize scope. Without it, agent messages show no name.Questions? Comment below or reach out to your Customer Success Manager.

What’s changing?Effective November 3, we’re adding two new fields, agent_name and agent_id, to the detail object returned by the /export/v1/records/interactions endpoint within the Moveworks Data API. These fields identify the assistant that handled each conversation, making it easier to break down usage, trends, and resolution data by Assistant.If all of your org’s interactions go through the Universal Assistant, agent_name returns “AI Assistant” If your org has no Productivity Assistants configured, agent_name returns “AI Assistant” by default (for the Universal Assistant). If a Productivity Assistant is configured (e.g., “Talent Insights”) for your org, interactions routed to it return that assistant’s name.What you need to know about the impact? This is a non-breaking change. The new fields are additive, and existing pipelines will continue to work as-is with no action required. To keep your Data API integration resilient to changes like this, please make sure it is built using the Data API integration best practices help doc.Note: There’s no change to row volume and no backfill. agent_name and agent_id will only be populated for interactions from November 3 onward. Resources:Interactions endpoint reference Data API release notes (agent_id and agent_name) Data API Integration best practices Data API: Breaking vs Non-Breaking changesIf you have any questions or concerns, feel free to comment on this post and our team will get back to you.Thanks,Moveworks Analytics team

You can now move from EmployeeWorks to Moveworks without signing in again. While signed in to ServiceNow, select Admin Tools → Moveworks Tools to access the Moveworks tools available to you. No separate Moveworks SSO setup is required. RequirementsServiceNow EmployeeWorks Web App v1.5.2 or later. The EmployeeWorks integration is configured and your user account has been imported into Moveworks The Content Delivery plugin (com.sn_content_delivery) active. Either System Administrator (admin), or Employee Center Admin (sn_hr_sp.esc_admin) plus either Content Manager (sn_cd.content_manager) or Content Admin (sn_cd.content_admin).Current limitationsAutomatic sign-in is currently supported only through EmployeeWorks → Admin Tools → Moveworks Tools, for users with the ServiceNow roles listed above. Other entry points, including the configuration-page Moveworks Setup link, continue to use the existing login flow.Refer to https://docs.moveworks.com/ai-assistant/employeeworks-web-app/access-moveworks-from-employeeworks for access steps and troubleshooting.

OpenAI is retiring an older set of LLM models. To keep our customers on a current, supported set of models, Moveworks is automatically upgrading any affected LLM action to a model of equal or better quality.What's changingOn Friday, 10/23, OpenAI is retiring a set of older models. Any Tool Studio LLM action in your org that also uses one of these models is being automatically upgraded to a supported replacement of equal or better quality. Customers who rely on the platform default rather than picking a specific model are not affected.Models that are retiring and their new replacements:gpt-3.5-turbo maps to gpt-4.1-mini gpt-4.1-nano maps to gpt-4.1-mini o1 maps to gpt-5.4 o3-mini maps to gpt-5.4-mini o4-mini maps to gpt-5.4-miniAll replacement models are available in every region.Who is affectedThe replacement models are available to all customers when setting up new LLM actions, while the retired models will no longer be available.Only customers with a Tool Studio LLM action that explicitly specifies one of the retiring models will experience an automatic update. If you rely on the platform default (i.e., leaving the model selection blank), no changes will occur, and you won't be affected by future model migrations either.What you need to doNothing. The change is applied automatically with no expected downtime. These replacement models run faster and they will return the same response structure.More information on LLM Actions: https://docs.moveworks.com/tool-studio/actions/llm-actionsLLM Action best practices: https://help.moveworks.com/tool-studio/guides/best-practices/llm-action-best-practices  

We're excited to share that Custom Views is now available for the View Feedback table in Moveworks Analytics.What's newMoveworks Analytics Admins can now create custom views from the “View Feedback” data and share curated, governance-friendly views with specific teams — giving stakeholders like ITSM admins, HR teams, or CS managers visibility into the feedback most relevant to them, with full control over what data is exposed. What you can doBuild a custom view from the View Feedback table, selecting only the columns you want to share (e.g., feedback type, topic, resolution status) while leaving out PII fields like user email and conversation ID Share the view directly with specific users Recipients only need the Custom Views User role — they never see the raw View Feedback table Who can create custom views from View Feedback?Any User with the Interactions Viewer role (found under Roles & Permissions → Bot Analytics) can access the View Feedback table and create custom views from it. How to get startedNavigate to Analytics → AI Assistant Insights → View Feedback Click on Create View Provide a name and description. Choose the columns that you want to remove from the view and share with specific users.  For more info, see our Custom Views Help Doc.For any questions, please feel free to reach out here, and our team will get back to you.  Thanks,Moveworks Analytics Team

Hi Moveworks Community, We’ve upgraded the infrastructure behind Triage Insights to a more modern, standardized data pipeline, making your Triage reporting more accurate and reliable. Triage in Moveworks applies only to ticket data, predicting and auto-filling fields like Category, Subcategory, Business Service, and Priority. Each ticket gets a Triage Result: Correct, Incorrect, No Prediction, or Not Relevant. Previously, for less than 1% of tickets, the system didn’t always pick the most accurate record, so some tickets showed the verdict as Not Relevant even when a correct verdict existed. That’s now fixed, so the pipeline always uses the latest data and each ticket’s verdict reflects its most recent triage. Note: No action is needed on your side. This improvement is automatic for all customers already using Triage on Moveworks. This is a forward fix and will only apply to all new triage runs starting 16th October.

The Moveworks Assistant in Slack is getting a faster, clearer experience! Answers stream in as they are written, employees can see what the Assistant is working on, tables display cleanly, and live agent chats are easier to follow.Most of these changes turn on automatically. A few are optional, because they depend on your preferences or need a change to your Slack app. This post covers what is changing, when, and what your admins can do.What's the timeline?October 19: The new experience goes live for a small group of early customers. From the same day, every other customer can try it in their sandbox. October 29: General availability. Every remaining customer using the Moveworks Assistant in Slack is upgraded automatically. You don't need to do anything to receive the automatic changes.What changes automaticallyReasoning steps: While the Assistant works on a request, employees see a short, collapsible plan of what it is doing instead of waiting on a blank reply.Tables: Answers that contain tabular data display as clean, native Slack tables.Rich text: Answers use Slack's native formatting, which makes longer answers easier to scan. Streaming answers: Answers appear as they are written, so employees start reading sooner.Optional features your admins can turn onThese stay off until an admin turns them on.Threading: Employees can keep separate conversations in separate threads in their DM with the Assistant, so context from one topic doesn't carry into another. To turn it on, go to Moveworks Setup > Chat Platforms > Manage Chatbots > Chatbots, edit your Slack bot, and select Allow Multiple Threads (Chats). The setting applies to everyone who uses that bot. Native pop-ups: Get Help, References, and feedback open inside Slack from a "..." menu on each answer, instead of sending employees to a web page. Feedback and Get Help move one click deeper, under the menu. To turn it on, go to Moveworks Setup > Chat Platforms > Manage Chatbots > Chatbots, edit your Slack bot, and select Enable Native Slack Modals. This allows for Get Help, Feedback, and References pages to be rendered in Slack without the need for a user to be redirected to their browser Work Objects: Tickets appear as live cards that employees can act on, for example to add a comment or resolve the ticket, without leaving Slack. This takes two steps. Your Slack admin updates the Moveworks app at api.slack.com/apps. Under Work Object Previews, turn previews on, add the entity type slack#/entities/item, and save. Under Event Subscriptions > Subscribe to bot events, add entity_details_requested and save. Then reinstall the app to your workspace. In Moveworks Setup > Chat Platforms > Manage Chatbots > Chatbots, edit your Slack bot and select Enable Work Objects. If your organization uses Channel Resolver:Channel Resolver: In support channels, your administrator will be able to choose the following on a per channel basis: Whether the Assistant responds to every message Which plugins it can use in that channel, including your own Tool Studio plugins. Whether replies from Moveworks appear in the channel or via DM Configuration: Each channel is set separately in Moveworks Setup > Core Platform > Channel Resolver > Channel Settings. When you add plugins to a channel's list, only the plugins on that list run in that channel, so add the built-in ones you want to keep as well. Good to knowFeedback and Get Help clicks are tracked the same way with native pop-ups on, so your existing reports keep working. The new experience is available in every data center, including GovCloud, at no additional cost. This update applies to the Moveworks Assistant in Slack only. Microsoft Teams and other chat platforms are not changing at this time.Questions?Reply in the comments below, or reach out to your Customer Success Manager.

We've released the Reducing Source-System Mentions in Assistant Answers model upgrade to Standard customers.When the Assistant answers from search results, it no longer names the system the content came from. Answers now focus on the content itself. Citations and source links work the same as before.For example, you should no longer see phrases like these in answers:"According to an article in ServiceNow..." “Based on the information I found in Content Gateway..."Read more about the release here.Learn more about our model upgrade program here.

What are we changing?Sources on citation surfaces stay in their original language — with translation on demand. Today, Moveworks Assistant automatically translates cited source in the references list into the user's preferred language. With this change, that becomes a choice instead of a default:Sources appear as written on citation surfaces. An English policy sources is displayed in English. A Spanish knowledge article sources is displayed in Spanish. The language of the source is shown through a language code. One click translates a source. A translate control appears on any source in a language other than the user's preferred language. One more click brings back the original. Users can switch between the two. Answers are unchanged. Moveworks Assistant still replies in the user's query language, exactly as it does today.For the best user experience, set your preferred language to match the language you use with Assistant. Scope of change:EmployeeWorks Web, Assistant and Search Moveworks Standalone Web, Assistant and Search Assistant bot on TeamsNote that Slack is not in scope for this change and will be updated at a later time.Illustrative mocks belowStandalone Moveworks Assistant - Inline ResponseEmployeeWorks Moveworks Assistant - References panelTeams Assistant Bot - References panelEnterprise Search App — Search ResultsEnterprise Search App — Citation Popup in SummaryEnterprise Search App — Citations List in Summary Who will benefit from this change?Customers with multilingual workforce with employees that have preferred language different from the language a source was written in.Why it matters:The citation matches the document. A cited source is the evidence behind an answer. When its title and snippet are translated without notice, employees click through to a document that looks different from what they were shown. Translation when employees want it, not by default. Multilingual employees may prefer to read the original or know what the original language was. Anyone who prefers their own language gets it in one click and can switch back to compare.What's the timeline?Controlled availability begins on November 2, with General Availability scheduled for November 16, 2026. If you’re interested in joining Controlled Availability, please contact your Customer Success Manager. 

What is EmployeeWorks Mobile?EmployeeWorks is ServiceNow's next-generation enterprise-wide AI front door. — now available as a native mobile experience inside the Now Mobile app. Employees can access the conversational interface powered by ServiceNow Otto (built on Moveworks), personalized canvas, service catalog, knowledge articles, tasking and requests, and other AI-powered experiences directly from their mobile device, no separate app required. As of the September release, EmployeeWorks Mobile is available in Controlled Availability (CA). A limited set of customers can configure, evaluate, and deploy the experience in production. Who is eligible for Controlled Availability?Controlled Availability (CA) is open to existing EmployeeWorks customers who meet the following requirements: Platform version: AP6 or BP0 (September 2026 release) EmployeeWorks Web App deployed in at least sub-production environment Now Mobile is already deployed to your users Intended use: sub-production or production deployment where CA limitations are acceptable Customers who do not meet the version requirements or whose use case requires features not yet in CA scope should plan for the GA release in December 2026 (Safe Harbor Applies). Does signing up give me immediate access?No. Submitting the interest form registers your intent to participate — it does not enable EmployeeWorks Mobile on your instance or guarantee access. The ServiceNow product team will review submissions and contact eligible customers with next steps and enablement documentation. How do I register my interest?Complete the Controlled Availability interest form linked in this post. The form takes approximately 5 minutes and collects information about your environment, use case, and team readiness. Submission does not guarantee access.Register Here

Hello,We are excited to let you know about an additive change to the Moveworks Data API, taking effect on 14 October 2026.This change applies to all Data API consumers by default. Nothing is required from you to receive it — we are sharing it in advance so you can review and handle it on your side before the date if it affects you. What is changing Live Agent handoffs will be recorded as a new resource type in data_api_plugin_resources. This is available for ServiceNow live-agent handoffs only at this stage.Field Value dim_resource_type RESOURCE_TYPE_HANDOFF resource_id The live-agent interaction identifier in your connected system. For ServiceNow, the sys_id of the interaction record data_resource_details JSON containing external_system (for example, SNOW) data_resource_details_b64 The same details in typed proto format, using a new HandoffResource type is_cited Always FALSE for these rows  conversation_id, interaction_id and plugin_call_id behave exactly as they do for all other resource types and can be used to join to data_api_conversations, data_api_interactions and data_api_plugin_calls. Why This lets you correlate assistant conversations with the live-agent session they were handed to, so you can analyse handoff outcomes using your own external system's records.What to check before 14 OctoberIf your pipelines filter to specific resource types — no impact. Add RESOURCE_TYPE_HANDOFF only if you want the new data. If your pipelines read all resource types — you will start receiving additional rows. Please confirm that rows of a new type will not affect ingestion, row counts or existing reports. If you decode data_resource_details_b64 — add support for the new HandoffResource type or skip rows where dim_resource_type = RESOURCE_TYPE_HANDOFF.  Current limitation ServiceNow only. Handoff resources are emitted for ServiceNow live-agent handoffs. Handoffs to other live-agent systems are not included yet. Timeline6 October 2026 — this notice 14 October 2026 — change takes effectIf you have any questions, please contact your Moveworks representative. 

Max Capacity now makes it easier to understand connector validation results and identify when user identity ingestion needs manual setup.Clearer Connector Validation resultsSelect View details in the connector setup step to review validation results for areas such as Content, User, Permission, and Group. The checks shown depend on the connector.Expand an error or warning to understand the issue. Inspect request and response details where available to help troubleshoot it. After correcting the issue, select Re-Validate to run the checks again.This example shows an expanded validation failure with request and response details.Save behavior is unchanged. Required validation failures still block Save. Optional warnings alone do not block Save; all other required checks and setup requirements must still be satisfied. Successful validation does not mean ingestion has completed or content is ready to search.Identity setup guidance before ingestionThe Start ingestion dialog now warns you when user identity ingestion needs manual configuration for the connector. This can happen when your organization has a joining key that does not resolve to an email address.The warning directs administrators to Review Identity to complete the setup.If you see the warning, select Review Identity, configure user identity ingestion manually, and map the joining key correctly for this connector.The warning does not block starting ingestion. However, starting ingestion does not resolve the Identity setup issue. Until Identity is configured correctly and permissions resolve, users cannot search content from that connector.Find these updates in Moveworks Setup > Search > Configure Search > Max Capacity. No action is required because of this update.Learn moreMax Capacity Setup Configure User Identity

Over the coming weeks, you'll notice that Agent Studio is being renamed to Tool Studio across the Moveworks product, documentation, website, Community, and Marketplace.What's changing? Only the name. Tool Studio is the same environment you already use to build, test, and manage the capabilities that power your Moveworks/EmployeeWorks agents. Your existing work, configurations, and integrations are unaffected.Why the change? The new name better reflects what you actually do in this space: create and manage the tools your agents use to get work done.When will I see it? The name will roll out surface by surface, starting with the in-product experience and Community, followed by the Marketplace, external website and Help documentation. We expect everything to read "Tool Studio" by early November.Where do I report? If you see Agent Studio references starting in November, report them to your Customer Success contact.A couple of things to note:Older materials like past release notes, announcements, and blog posts will keep their original "Agent Studio" wording to preserve historical accuracy. If you spot "Agent Studio" somewhere after the rollout, no need to worry, it won't affect how anything works.

How are we improving the Assistant’s behavior?We’re improving how the Assistant handles lists of results during complex, multi-step requests. Previously, some requests could return an incomplete set of similar results when information was passed between steps within a tool execution.With this update, the Assistant can preserve and use more of the relevant results, producing more complete and consistent outcomes.Key datesFrontier: September 23, 2026 Standard: September 24, 2026 Basic: September 28, 2026Technical restrictionsNo customer configuration changes or actions are required. Availability follows each customer’s Minor ML Upgrade rollout tier.DocumentationNo additional customer documentation is required for this behavior improvement.

Your Moveworks AI Assistant can turn structured data into trusted answers in seconds. Many teams are already using Structured Data Analysis (SDA) today. Now we're putting the control directly in your hands: a setting to enable or disable SDA now appears in your Moveworks Setup.To make sure every team can take advantage of it, SDA will be enabled for your organization on October the 14th unless you choose to turn it off.Why you will love itWith SDA, employees just ask in plain English and get a real answer back in seconds, because the Assistant writes and runs actual code against the data, then shows exactly what it did.Faster insights. Ask in everyday language and get a ranked, calculated answer instead of a guess. Real analysis, not truncation. The Assistant works across the full dataset, aggregating, filtering, and ranking, instead of losing accuracy to token limits. Answers you can trust. Every response includes a plain-English summary plus an expandable code citation showing the exact code, variables, and outputs. No black boxes. Safe by design. All code runs in an isolated, resource-capped sandbox that's wiped after each run.What this means for you The choice is now yours. The SDA setting lives in your Moveworks Setup, under AI Assistant > Plugin Management > Code Interpreter Plugin > User filter for enabling plugin, so you can turn it on or off whenever you like. Already using SDA? Nothing changes. It stays on. Already have SDA turned off? Nothing changes. It stays off. Haven't set a preference yet? SDA will be turned on for your organization on 10/14 so your team doesn't miss out.   If your filter is unset, it will be changed to TRUE on 10/14. What you need to do            Make sure your setting matches your wishes.Is it turned on in the settings and you want to keep it on? Nothing at all. It'll be active for your team on 10/14. Prefer to hold off? Just set the user filter for Code Interpreter Plugin to FALSE on your Plugin Management administration page, or use user.email_addr IN [...] to gate it to one or more testers, and you can turn it back on anytime.Read more about SDA here and here. 

We’re excited to roll out a major upgrade to how Moveworks ingests content from your connected systems. This upgrade helps us download large amounts of content faster and respond to changes in content filters without downloading everything again. It also lays the foundation for the upcoming Content Troubleshooting Hub.Moveworks will manage the rollout internally. You do not need to migrate connectors or change your configuration.What improvesDownload large amounts of content faster: Moveworks can download large amounts of content more quickly, reducing the time spent bringing it in from your connected systems. Change content filters without downloading everything again: When content filters change, Moveworks rechecks content it has already found against those filters every 15 minutes, without repeating the entire discovery and download process. Prioritize recently updated documents during full refreshes: Moveworks downloads recently updated documents first when performing a full refresh.What to expect during the rolloutWe are rolling out the upgrade connector by connector, with completion planned by October 31, 2026. For the connectors in this rollout, we ran the updated ingestion process alongside the existing process and verified that the ingested content and permission records matched.What changes during the rolloutAs each connector is upgraded, you’ll be able to investigate connector issues via failed API requests directly, instead of first opening an ingestion run.Upgraded connectors appear under Content (new architecture) on the standalone Setup → Core Platform → Data Ingestion → Data Crawling page. Other connectors remain under Content until they are upgraded, so both sections can appear during the rollout.The crawling dialogs inside Indexed Content do not show this new section. Open the standalone Data Crawling page for failed requests, and continue using Indexed Content for individual record details.Crawl Health displays Not applicable because this section does not report an overall crawl-health status. The label itself does not indicate an ingestion failure.Upgraded connectors appear under Content (new architecture). Select View Failed Logs to investigate failed requests.Select View Failed Logs for the relevant connector and resource to see failed API requests from the last 30 days. You can filter the list by Error Code.The count represents failed API requests, not missing documents. Request URLs are hidden in this example.See the Data Crawling guide for failed-request troubleshooting and the Indexed Content guide for individual record details.Coming next: Investigate content problems in one placeDuring this transition, you’ll use Data Crawling for failed requests and Indexed Content for individual record details. We’re developing Content Troubleshooting Hub, planned by November 30, 2026, to bring these details together with guidance on what to do to resolve these active issues. 

We've released the Clearer Outcomes for Tool Calls model upgrade to Basic customers. How are we improving the Assistant’s behavior?Failed or empty tool calls now show a clear failure or no-results state, along with guidance on what the user can do next, instead of appearing successful.When is the release date?Frontier: September 9, 2026 Standard: September 14, 2026 Basic: September 17, 2026Technical RestrictionsThis initial release primarily covers Agent Studio Plugins. MCP failure transparency, and more specific action-error messages are not included in this release and are being handled as separate follow-up work.DocumentationDocumentation describing reasoning and tool-failure behavior will be linked here when published.Read more about the release here.Learn more about our model upgrade program here.

ToolSmith enters Controlled Availability on September 22, 2026, bringing a conversational way to build and maintain plugins in Tool Studio.Describe the workflow you want, review the proposed plan, and let ToolSmith help create and test the pieces behind it. You can also bring an existing plugin into the conversation to make a change or investigate a failure.What you can do with ToolSmithBuild a new plugin from a description. Explain the outcome and the systems involved. ToolSmith researches available APIs and existing assets, proposes a plan, and builds after you approve it. Edit an existing plugin. Describe what needs to change, review the proposed approach, and work through the update and testing in the same conversation. Debug a failing plugin. Provide the issue and relevant conversation context. ToolSmith can inspect execution logs, investigate the cause, and propose a fix.For example, start with: “Build a plugin that shows my open Linear issues.” Or bring an existing plugin and ask: “Add a priority filter.”Review the plan. Stay in control.You can refine the plan in plain language before approving it. When credentials are needed, ToolSmith collects them through a secure configuration card. It also asks for approval before taking external actions, including tests that change data in connected systems.Your conversation keeps the plan, progress, and results, so you can return and continue where you left off.Join Controlled AvailabilityWe’re starting with a selected group of customers who have a real workflow to build, edit, or debug and a participating developer with access to Tool Studio and the relevant third-party systems.Interested? Request early access to ToolSmith by telling us about your team, the workflow you’d like to build or improve, and the systems it uses. Our team will review your nomination and coordinate next steps with your account team. Submitting a request does not guarantee access. Selected participants will receive a kickoff, guidance on their first use case, and ongoing opportunities to share feedback.During CA, start with a focused workflow involving one or two systems. Results vary by system and workflow complexity, so review and test generated changes before relying on them. When editing shared assets, check their dependencies in Tool Studio because ToolSmith does not display those dependencies in the conversation.Want to experience ToolSmith live? Join us at now[dev]26 in San Jose for two days of hands-on building with engineers and product teams.What would you like to build—or improve—with ToolSmith? Tell us in the replies.

Connecting the AI Assistant to your apps is now smoother. When a request needs access tosomething like Google Calendar, Zoom, or GitHub, the assistant handles it up front insteadof failing partway through.Pre-flight ChecksThe assistant identifies every connection a request needs before it runs, and asks for allof them at once. If a request needs both Google Calendar and Gmail, you approve both in asingle prompt, each with a clear description of what it allows.Auto-RestartOnce you approve, your original request continues automatically with all prior contextintact. No repeating yourself, no typing "retry." Step away mid-approval and you have a30 minute window to come back and pick up where you left off.What you need to doNothing. These features have been enabled by default across all orgs. No configuration or admin actionrequired. 

We've released the Clearer Outcomes for Tool Calls model upgrade to Standard customers. How are we improving the Assistant’s behavior?Failed or empty tool calls now show a clear failure or no-results state, along with guidance on what the user can do next, instead of appearing successful.When is the release date?Frontier: September 9, 2026 Standard: September 14, 2026 Basic: September 17, 2026Technical RestrictionsThis initial release primarily covers Agent Studio Plugins. MCP failure transparency, and more specific action-error messages are not included in this release and are being handled as separate follow-up work.DocumentationDocumentation describing reasoning and tool-failure behavior will be linked here when published. Read more about the release here.Learn more about our model upgrade program here. 

Clearer Outcomes for Tool Calls is now available to Frontier customers. When a tool call fails, times out, or returns no results, users will now see a clearer explanation of what happened and, where applicable, guidance on what to do next. This makes unsuccessful tool calls easier to understand and reduces ambiguous or incomplete assistant responses. Technical RestrictionsThis initial release primarily covers Agent Studio plugins. MCP failure transparency and more specific action-error messages are not included in this release and are being handled as separate follow-up work. Future Release Dates:Standard: September 14, 2026 Basic: September 17, 2026

We've released the Context-aware toxicity filtering model upgrade to Standard customers.Key DatesFrontier: September 2 2026Standard: September 8 2026Basic: September 10 2026Learn more about our model upgrades here.We’re upgrading the AI Assistant’s toxicity filtering to consider the full context of a conversation rather than evaluating content in isolation. This provides a more accurate understanding of user intent and helps identify genuinely harmful content. The upgrade also reduces false positives, allowing safe conversations to continue without unnecessary interruption. Administrators can review the configuration steps for more information: https://help.moveworks.com/agent-studio/agentic-ai/conversational-safeguards-assistant/configure-moveworks-toxcity-filter#configuration-stepsRead more about the release here.Learn more about our model upgrade program here.

 🔄 What's changingLong documents, such as a detailed policy or a multi-step procedure, often hold an answer across more than one section: a prerequisite in one place and the steps that follow it in the next. Today, Moveworks answers from the passage that best matches a question, so an answer that spans multiple sections of a long document sometimes comes back incomplete.Two things happen when someone asks a question, and they want different things from a passage:Finding works best when passages are small and precise. The tighter the passage, the more accurately it matches a question. Answering wants the opposite. More surrounding context makes the answer more comprehensive.Until now, one passage did both jobs. Going forward, we will separate them: search still runs on the same small, precise passages, and once the right one is found, Moveworks assembles a wider window around it — the neighboring passages from the same document — and hands the Assistant that fuller context. When several passages from one document rank highly, Moveworks treats it as a strong source and pulls in more of it.✨ What this improvesLong procedural articles: Steps and their prerequisites often sit in different sections, and the Assistant now sees both of them together, so the answer covers the whole procedure rather than one part of it. Policy documents with nested headings: A subsection is read alongside its neighboring sections and its document title, so the meaning it inherits from the material around it carries through to the answer. Questions that span multiple sections: An eligibility rule and the process that follows it are returned as a single complete answer instead of two separate results.✅ What you need to doThere is no configuration needed from the admin or user. This update will be rolled out to all customers by 9/30/26. Refer to our roadmap: here.  

We’re upgrading the AI Assistant’s toxicity filtering to consider the full context of a conversation rather than evaluating content in isolation. This provides a more accurate understanding of user intent and helps identify genuinely harmful content.The upgrade also reduces false positives, allowing safe conversations to continue without unnecessary interruption.Example OutputContent safety is now handled directly by the Assistant's reasoning engine, which evaluates each request in the full context of the conversation. Instead of a context-blind checkbox, the Assistant understands what you're actually asking for before deciding whether it can help. When a request genuinely crosses a line, the Assistant issues a clear, consistent refusal.The old "Specific Topic Toxicity" model, with its fixed five fixed categories, has been replaced by a free-form custom topic prompt. In the new model, six categories remain blocked by default in the background, and you can now add your own on top of them.Default blocked categoriesThese six categories are blocked by default for every organization:Violent — physical violence Non-violent Illegal Acts — requests to carry out unlawful activity that isn't violent Sexual Content — sexually explicit content Unethical Acts — content promoting unethical behavior Politically Sensitive — contentious political content Jailbreak — attempts to bypass the Assistant's safety instructions Note: Core anti-jailbreak protections always remain in place and cannot be overridden. Rollout Schedule/ TiersFrontier: 9/2/2026Standard: 9/8/2026Basic: 9/10/2026 What improvedFar fewer wrongly-blocked requests. The old filter frequently flagged benign messages as unsafe. Across 5,000+ live production conversations, the new approach produced zero user-visible false refusals, so legitimate requests get answered. Better at catching genuinely unsafe content. On calibrated safety benchmarks, the new approach catches roughly twice as much truly harmful content as the previous filter. Context-aware judgment. The Assistant can now tell the difference between, for example, "a colleague is harassing me, how do I report it?" (helped) and harassing content itself (refused). It also holds up regardless of framing, whether fiction, roleplay, hypotheticals, or jokes. Broader, more sensible coverage. Requests like "help me download movies and software without paying" are now correctly declined. New: per-organization safety customization. Enhanced content-safety control for your organization.Every other aspect of the Assistant's behavior is unchanged. This only affects when and how genuinely unsafe requests are declined.How to check it out in your orgSee it in action. No setup required. The upgraded safety behavior is enabled for your organization automatically as part of our staged rollout. You can simply keep using the AI Assistant, and legitimate requests that may have been over-blocked before should now go through cleanly.Customize safety rules for your organization. A new Additional Organization Safety Instructions setting lets you extend the built-in safety categories with your own rules. For example:Find in: Chat Platforms -> Display Configurations > Moveworks AI Assistant Display Settings & Disclaimers > Safety Guard: Allowed Content Categories New AdministrationOld Administration (Replaced)Update Settings if Needed: This free-text setting replaces the previous category allow-list. If your org previously configured the content-safety categories, you will want to customize the new instructions accordingly.Adding topics your organization wants the Assistant to decline, or Defining approved exceptions (for example, authorizing your security team to request phishing-simulation drafts for awareness training).More information in the help documents here.