Grok Bot 1.0 and the November 11 update: xAI's teammates get meetings, Teams, and an admin layer
xAI launched Grok Bot 1.0 on November 2 with shareable Team Bots, proactive work pickup, and faster browser use, then shipped a November 11 update adding Google Meet, Zoom and Microsoft Teams meetings, shared connectors, and a Review learnings approval queue. What shipped, how the permission model works, and what to test before rolling it out.
Grok Bot 1.0 launched on November 2, taking xAI's AI teammate out of beta with three changes: Bots you can share with a whole team across Slack, Microsoft Teams and video calls, Bots that spot and offer work without being asked, and faster, more reliable browser and document work. A follow-up update on November 11 (v1.0.1) added the enterprise plumbing: Bots can join Google Meet, Zoom and Teams calls from a link, Team Bots now work inside Microsoft Teams, owners can share app logins as shared connectors, and anything teammates teach a Bot waits in a Review learnings queue until an owner or Manager approves it. The launch details matter less than the permission model underneath them, which is the part worth studying if you build or buy agents.
What Grok Bot 1.0 actually shipped
Per xAI's launch post, the 1.0 release rests on three pillars:
- Multiplayer teammates. Share a Team Bot with specific people or groups, and work with it in Slack or Microsoft Teams. Bots follow multi-person discussions, understand who is involved, and contribute or offer to pick up work. They can also join video calls through the chat.
- Proactive teammates. Bots monitor ongoing projects, recognize when something needs attention, follow up on unfinished tasks, and bring people in when a decision needs human judgment. xAI's example: a Bot notices a launch deadline approaching and prepares the materials ahead of time.
- More capable teammates. xAI claims better browser navigation and sign-in handling, stronger documents and slide decks, interactive Pages you can edit together, improved Gmail, Google Calendar and GitHub connections, and lower resource use per task.
What the November 11 update adds
The v1.0.1 changelog entry, titled "Meetings, Microsoft Teams, and shared connectors", is where the release becomes deployable for companies:
- A Bot can join a Google Meet, Zoom or Microsoft Teams call from a link, answer questions in the meeting chat, and send a summary afterward (Team, Enterprise and Ultra plans).
- Team Bots join Microsoft Teams through a three-step approval: a team admin turns it on, an owner or Manager connects the Bot, and a Teams admin approves it.
- Owners and Managers can share their own app login as a shared connector, but only after a team admin enables Allow Shared Logins.
- What teammates teach a Team Bot no longer applies immediately: it waits under Review learnings until the owner or a Manager approves it.
- Team admins can limit who is allowed to share Bots at all, through a Bot sharing control.
The admin and permissions layer is the real story
Most agent launches lead with capability. The interesting part of this one is how xAI scoped authority, because it mirrors the problems anyone deploying shared agents hits:
- Learning is gated, not automatic. A Bot that learns from everyone it works with is also a Bot that can learn the wrong thing from anyone. The Review learnings queue makes knowledge changes an explicit approval, the same pattern you want for any agent with persistent memory.
- Credential sharing is an admin decision. Shared connectors let one person's app login serve the whole team, which is convenient and dangerous in equal measure; xAI gates it behind an explicit admin toggle (Allow Shared Logins) rather than making it a default.
- Sensitive actions stay human-approved. Per the launch post, Bots respect each person's permissions, ask before sensitive actions, and check whether information is appropriate for a public channel. xAI's example: asked to share a deck with an external agency on a call, the Bot flags that slide 6 has sensitive financials and offers to send it without that slide.
- Recent changelog entries show the same thinking elsewhere: commands that open or drive a Tesla always ask for approval, auto-review cards say which rule paused an action, and a revoked X Money connection stays disconnected until the user reconnects it in the app.
None of this is independently verified; it is how xAI describes its own controls. But as a design checklist for shared-agent permissions, it is a solid reference.
Availability and plans
Per the launch post: Team Bots are available on Team and Enterprise plans. Meetings are available on Team, Enterprise and Ultra plans, and Enterprise admins must turn on Meetings in settings. Grok Bot 1.0 is available now for creating a Bot, sharing it, or bringing it into Slack, Microsoft Teams and video calls.
What I would check
My untested suggestions before rolling this out to a team:
- Test the Review learnings workflow with real teammates before trusting a Bot's accumulated knowledge; check what happens when two people teach conflicting things.
- Audit what a shared connector actually exposes: whose data, which scopes, and what the Bot can do in Slack or Teams channels under a shared login.
- Run a mock meeting with sensitive content and verify what the Bot puts in the chat and the afterward summary, and who receives it.
- Read the changelog's fixed list as a maturity signal: this is a young product with many rough edges patched weekly, so pilot with a low-stakes team first.
- Compare the permission model against what you already get from Claude or Microsoft Copilot deployments before switching.
What I would not conclude
Do not conclude the learning and proactivity claims work as described; they are xAI's own, and a proactive Bot is also one that can act on a wrong understanding, which is why the approval queues matter. Do not assume feature parity outside X's ecosystem; the launch leans on Slack, Teams, Google and GitHub connectors, and depth elsewhere is unproven. Do not treat "out of beta" as "enterprise-proven"; the changelog's pace of fixes suggests otherwise. And do not read the safeguard descriptions as a guarantee; test the controls on your own data before giving a Bot real responsibility.
Frequently asked questions
What is Grok Bot 1.0?
xAI's AI teammate product, moved out of beta on November 2, 2026. Bots can be shared with a team across Slack, Microsoft Teams and video calls, can proactively spot and offer work, and got faster and more reliable at browser and document tasks, according to xAI.
What did the November 11 Grok Bot update add?
Per xAI's changelog, v1.0.1 lets a Bot join Google Meet, Zoom and Microsoft Teams calls from a link with meeting-chat answers and summaries, brings Team Bots into Microsoft Teams, adds admin-gated shared connectors for app logins, and requires owner or Manager approval for anything teammates teach a Bot.
What are Team Bots?
A Team Bot is one AI teammate shared by a group: everyone works with the same Bot, and it builds understanding of the team's work, preferences and decisions over time. Sharing can be limited to specific people or groups, and team admins can restrict who may share Bots at all.
How does the Review learnings queue work?
When teammates teach a Team Bot something, the change does not apply immediately. It waits in a Review learnings queue until the Bot's owner or a Manager approves it, so one person's bad instruction does not silently become team-wide behavior.
Which plans include Grok Bot meetings and Team Bots?
Per xAI: Team Bots are on Team and Enterprise plans. Meetings are on Team, Enterprise and Ultra plans, and Enterprise admins must enable Meetings in settings first.
Can a Grok Bot share my app logins with my team?
Owners and Managers can share their own app login as a shared connector, but only after a team admin turns on Allow Shared Logins. It is an explicit opt-in at both levels, not a default.
Are xAI's safety and permission claims verified?
No. The descriptions of approval behavior, permission respect and public-channel checks come from xAI's own launch post and changelog. They are a design checklist worth testing on your own data, not an independent audit.
Sources
- Introducing Grok Bot 1.0 | xAI - November 2, 2026, read November 12, 2026. The three pillars (multiplayer, proactive, more capable), Team Bot sharing, video-call participation, safeguard descriptions, the sensitive-financials approval example, and plan availability.
- Grok Bot Changelog | xAI - v1.0.1, November 11, 2026, read November 12, 2026. Meetings on Meet, Zoom and Teams, Microsoft Teams integration approval flow, shared connectors and Allow Shared Logins, Review learnings, Bot sharing admin control, plus the November 5-6 entries on approval rules and connection behavior.
- TLDR AI newsletter, November 12, 2026 - the tip that surfaced the update.
Not read: xAI's plan pricing pages, admin documentation, and security whitepapers beyond the two pages above. All capability, learning and safeguard descriptions are xAI's own claims; I did not test the product.
Related reading
- Self-hosted Claude Managed Agents: Anthropic's agent service now runs on your infrastructure
Anthropic opened self-hosted Claude Managed Agents in private beta: you host the agent service, its database, and the sandboxes, while model calls go to the Claude API or a provider you choose. What runs where, which credentials exist, what still leaves your network, and what to check before committing.
- Anthropic's always-on agents guide: the five primitives behind agents that never sit idle
Anthropic published a guide on building always-on agents: stateful, proactive agents that manage context across sessions. The core idea is that the agent is not actually running all the time - it sleeps between runs, and five architectural primitives decide when it wakes, what it remembers, and what reaches you.
- Semantic Kernel Python 1.46.0: Tool Authorization Docs, Template Trust and Pinned DNS in HttpPlugin
Semantic Kernel Python 1.46.0 mixes security fixes with documentation that says out loud what its function filters do not do. I read the release notes, the pull requests and the README at the tag to see what changed and what still falls on your application.
The build log
Follow the work.
New essays and prototypes, in your feed reader. Email subscriptions are paused while I connect a working delivery service. No email is collected here.
Follow via RSS