Skip to main content

Coming soon

Last updated: July 15, 2026

New since last update​

End user features​

Assistant​

  • GLM 5.2 in Assistant model picker via Fireworks (Glean key only) — Glean Assistant will soon support GLM 5.2 as a selectable model in the Assistant model picker for Glean key customers using Fireworks as the provider. Users will be able to experiment with GLM 5.2 on their own company data within Glean's governed environment, alongside other leading models, to compare quality, latency, and behavior. This launch will be limited to the model picker surface only and will not change an organization's default Assistant model. Nemotron 3 Ultra will not be included in this release. (Glean ID: ROAD-1576-MR)
  • GitHub sync and import for skills (single skills) — Admins will soon be able to import and sync skills from public GitHub repositories into Glean. This will make it easier to manage shared skill libraries in Git and keep Glean up to date with community and open-source skills that follow the open agent skills standard. Users will be able to paste a public GitHub repo URL, and Glean will fetch compatible .skill and .md files for import. Imported repositories will stay up to date through an hourly sync, and admins will be able to view sync status or disable syncing for a repo at any time. (Glean ID: ROAD-1407)
  • Automatically refresh interactive artifacts on a schedule — Scheduled refreshes will soon keep interactive artifacts up to date automatically. Artifact creators will be able to choose a daily, weekday, weekly, or monthly schedule, manage it from the artifact or active agents, and run a refresh at any time. (Glean ID: ROAD-1574)
    Library refresh: Find and revisit artifacts by owner, team, and popularity
  • Favorite artifacts to find them faster in the library — Users will soon be able to favorite artifacts from the library or directly in canvas. Favorited artifacts will appear at the top of the library in the Favorites section, making them easier to find and revisit. (Glean ID: ROAD-1567)
  • Engram: agentic memory system for Glean Assistant — Engram will soon become Glean Assistant's new agentic memory system that builds a durable, hierarchical view of each user's work across chat, documents, tickets, and other activity so Assistant can personalize responses, continue long-running projects, and reduce repeated context-setting. Memory will be persistent, privacy-safe, and scoped to each deployment and user, so Glean will never learn across companies and will only use content users already have permission to access. Users will be able to inspect, edit, or delete what Glean remembers in a dedicated memory view, giving them direct control over how the assistant learns from their role, preferences, and projects. Engram will initially be available for Glean Assistant end users on GCP deployments with Glean Universal Key, with support for additional cloud environments following in future releases. (Glean ID: ROAD-1412)

Agents​

  • Agent identity: agents act and appear as themselves via scoped service credentials — Glean agents will soon be able to run with service credentials instead of borrowing the invoking user's. Admins will be able to register service credentials, such as a Slack bot, Atlassian account, GitHub App, or GCP service account, and scope them to specific tools, while builders attach them to an agent. The agent will then act as itself in connected systems, with the triggering user or schedule recorded in Glean's audit log. (Glean ID: ROAD-1542)
  • Agents consumption UX revamp: trigger-first library and post-publish flow — Glean will soon make it easier to turn on scheduled and content-triggered agents by introducing a trigger-first view in the agent library. When users open an eligible agent, they will see its trigger conditions, schedule, and activation controls up front, so it is clear what the agent does and how to start it. After publishing one of these agents from the builder, Glean will also redirect users to this trigger view so they can confirm setup and activate it without extra clicks. This experience will ship together with Content Triggers GA and will be available to users where agents and triggers are enabled. (Glean ID: ROAD-1524)

Connectors​

  • Expand code search, code writer product additions, and support GitHub Enterprise and GitLab — Glean will soon add support for GitLab and GitHub Enterprise for code writers. (Glean ID: ROAD-1124)

Embedded integrations​

  • Run Glean agents from the Slack sidebar — Users will soon be able to run Glean agents directly from the Glean sidebar in Slack DMs, without switching to the Glean web app. From within Slack, users will be able to pick an agent, provide inputs, and see results inline in the sidebar, keeping IT requests, workflows, and everyday automation in one place. This will make it easier for teams who primarily work in Slack to get agent-powered help and complete tasks faster, end to end. (Glean ID: ROAD-865)

Glean Home​

  • Improved user profile menu — The user profile menu in the new left navigation will soon make common actions easier to find. From the profile menu, users will have quick access to install Glean apps, customize their theme, and open Help to learn more about Glean or give feedback. This change will affect end users on the new left navigation and will replace the Help button on the Glean home page. (Glean ID: ROAD-1563)

Admin features​

Admin capabilities​

  • Granular per-user usage breakdown in usage dashboard — Admins and end users will soon be able to see a granular breakdown of Glean usage at the individual level in the Usage dashboard. For Enterprise Flex and Glean Core Suite customers, admins will be able to use the By users view to understand which product types, models, and agents are driving a specific person's usage, making it easier to investigate spikes and manage spend. End users will be able to see a similar breakdown for their own account in user settings, so they can understand what is contributing to their monthly usage and adjust their behavior if needed. (Glean ID: ROAD-1568)
  • Sortable and exportable by users table in usage dashboard — Admins on Enterprise Flex and Glean Core Suite will soon be able to sort and export the By users table in the Usage dashboard, making it easier to analyze who is driving AI usage. The table will support server-side sorting on month-to-date usage and usage limits, department filtering, search, and fast pagination even for very large organizations. Admins will also be able to export the full By users view to CSV for deeper analysis, chargeback workflows, or internal reporting without slowing down the in-product experience. (Glean ID: ROAD-1569)

Connectors​

  • Using severity analysis to introduce a new category "False positive" to reduce noise in "Low" — Glean's sensitive content models will soon use policy-level custom instructions to automatically suppress known false-positive patterns instead of turning every low-severity match into a new issue. Admins will be able to mark categories of findings as false positives so they no longer create issues, while severity analysis continues to highlight truly risky documents. This will reduce review noise, improve trust in sensitive findings, and make it easier to adopt continuous scanning without overwhelming moderators. (Glean ID: ROAD-1493)

Protect​

  • Using severity analysis introduce a new category "False positive" to reduce noise in "Low" — Glean's sensitive content models will soon use policy-level custom instructions to automatically suppress known false-positive patterns instead of turning every low-severity match into a new issue. Admins will be able to mark categories of findings as false positives so they no longer create issues, while severity analysis continues to highlight truly risky documents. This will reduce review noise, improve trust in sensitive findings, and make it easier to adopt continuous scanning without overwhelming moderators. (Glean ID: ROAD-1493)
  • Export guardrails violations via UI and API — Admins will soon be able to export filtered AI security violations from Glean Protect for audit, reporting, spreadsheet review, and downstream security workflows. Exports will support the same filters used in the violations dashboard, making it easier to move findings into external review and SIEM-style processes without manual copy-paste. This will be for admins using Glean Protect AI security. (Glean ID: ROAD-1337)
  • AI security partner integration API — Glean Protect+ will soon support AI security partner integrations, giving admins more flexibility to route AI traffic through their preferred security provider instead of relying only on Glean’s built-in guardrails or a single partner option. This will help organizations apply consistent AI security controls across Glean Assistant and agents while continuing to detect and manage threats such as prompt injection, malicious code, and harmful content through centralized policies and findings workflows. (Glean ID: ROAD-974)

Still coming soon​

End user features​

Assistant​

  • Assistant typography refresh and visual polish — Assistant will soon use Inter typography with updated spacing and visual hierarchy for a more consistent, easier-to-read experience across all devices. (Glean ID: ROAD-1526)
  • Create podcast and audio artifacts with Assistant — Users will soon be able to create podcast and audio artifacts in Glean Assistant. By asking for a podcast on any topic, they will get a playable audio artifact they can listen to, share with teammates, and iterate on. (Glean ID: ROAD-1373)

Connectors​

  • Glean Assistant on external AI platforms via A2A — A new cross-platform A2A foundation will soon let Glean Assistant run as a governed, first-class agent inside external AI platforms and agent runtimes, starting with Gemini and expanding to other ecosystems over time. The shared protocol and packaging model will reuse Glean’s enterprise context, permissions, and guardrails so users can ask Glean from their preferred AI surface without duplicating configuration or exposing extra data. Admins will get a consistent way to authorize external platforms, manage which Glean capabilities are available where, and monitor usage across environments. (Glean ID: ROAD-1463)
  • Simplified MCP OAuth consent and scope surface — Connecting third-party MCP hosts to Glean will soon use a single, clear consent screen instead of a long list of technical OAuth scopes. The new experience will group MCP access into one user-facing permission, align that approval with each person’s existing permissions and admin-approved settings, and narrow the underlying scopes to only the resources the host needs. Additional instrumentation on consent approvals and declines will help admins monitor rollout safety and refine their MCP strategy over time. (Glean ID: ROAD-1448)
  • Glean in Gemini agent marketplace — Glean will soon be available in the Gemini agent marketplace, bringing the Glean Assistant into Gemini agents. (Glean ID: ROAD-1449)

Embedded integrations​

  • Glean in Microsoft Copilot — Glean will soon be available in Microsoft Copilot Studio, Copilot Chat, and Teams, bringing Glean’s enterprise knowledge directly into Microsoft’s AI surfaces. (Glean ID: ROAD-1112)
  • Get trusted Glean answers directly in Slack channels — Glean will soon be able to post high-quality answers directly into Slack channel threads so everyone can see them immediately without first clicking View response. These public replies will be generated only from content that is broadly visible in the company, such as public Slack channels and organization-wide documents, so they will be safe to share and easy to trust. Teammates will still be able to personalize the answer with their own private context when they need a more tailored response. (Glean ID: ROAD-908)

Admin features​

Protect​

  • Sensitive content policies at runtime: user input — Sensitive content policies will soon give admins a consistent way to define and enforce protections for sensitive content across Glean. Today, Glean Protect+ applies these policies to indexed data, and this upcoming update will extend that protection to runtime interactions so customers can block sensitive content in live AI prompts and responses as well. This will help customers apply sensitive data policies more consistently across both stored content and in-the-moment interactions. (Glean ID: ROAD-1212)
Was this page helpful?
Last updated on
© 2026, Glean Technologies, Inc.