xAI Releases Grok Bot 1.0: One AI Bot for the Whole Team in Slack, Teams, and Video Calls
Most AI assistants are personal: you type a request, and your team never hears about it. Grok Bot 1.0 from xAI puts one bot to work with several people at once in Slack, Microsoft Teams, and video calls, while it remembers decisions and tracks deadlines. For example, Research Bot analyzes customer feedback after a launch and passes its findings to Copy Bot, which revises the campaign copy once the team approves. For product managers, this is a step from a chatbot toward a shared member of the project. Team Bots are included in the Team and Enterprise plans. Joining video meetings is also available on the Ultra plan. The bots respect access permissions and ask for confirmation before critical actions.
AI-processed from Habr AI; edited by Hamidun News
Most AI assistants still work like a personal notebook: you type a request, get an answer, and your team knows nothing about it. If five people are discussing a product launch, each of them has to separately explain the context to the bot. xAI (the source calls it SpaceXAI) has released Grok Bot 1.0, a bot system in which a single assistant works with several people at once in the workplace messengers Slack and Microsoft Teams, as well as on video calls. It remembers decisions made, keeps track of deadlines, and suggests what to do next. Here is how it works and what can be used right now.
One bot for the whole team
The main novelty is Team Bots. One assistant is connected to Slack, Microsoft Teams, and video conferences. It follows discussions involving several people and understands the context of the debate. The bot remembers decisions, takes into account the preferences of individual employees, and over time gets better at understanding the specifics of a particular team.
Imagine a typical product launch. Research Bot collects and analyzes customer reviews, then passes its conclusions to Copy Bot. Copy Bot adjusts the advertising campaign texts, but only after the team has approved the chosen direction. The decision stays with a human, while routine data handoffs between stages go to the bots.
In video meetings, Grok Bot can surface information from past discussions, add important details, and record the outcomes. Instead of an employee who takes minutes and then sends out a list of agreements, the assistant does this.
Initiative instead of waiting for commands
The second feature is proactivity. The bot tracks projects, monitors approaching deadlines, checks task progress, and points out what to focus on first. According to xAI, it can prepare the necessary materials before anyone asks for them. If a decision requires a person, the bot brings the right employees into the process itself.
This raises an immediate question: what happens if the bot shows too much initiative? xAI says it has built in protections. The bots take users' access rights into account, ask for confirmation before critically important actions, and check whether particular information may be published in open channels. This is a reasonable minimum: otherwise, an assistant that had seen a private discussion could accidentally post its contents in a shared chat. How well these mechanisms work in practice will only become clear from real deployments; there are no independent checks in the source.
What was improved and what it costs
Beyond teamwork, version 1.0 also upgrades the agents' everyday skills: browser navigation, document creation, and presentation preparation. The bots now log into websites more reliably and fail less often on complex tasks. It is now possible to create interactive pages that the team edits together. Integrations with Gmail, Google Calendar, and GitHub have been expanded.
Access is split by plan. Team Bots are included in the Team and Enterprise plans, while video meeting participation is also available on the Ultra plan. Individual users on basic plans will not get these features, according to the description.
For context: the company previously introduced its flagship model, Grok 4.7. In xAI's own tests, it scored 46.3% on CursorBench 4.0, versus 40.4% for Grok 4.6 and 41.7% for GPT-5.6 Sol. On DeepSWE v1.1 it scored 71% (GPT-5.6 Sol: 72.7%, Fable 5.1: 70%), and on Terminal-Bench 4.0 it rose from 20.3% to 38%. These are the company's internal measurements, and they should be treated with caution.
What practitioners should do with this
If you are building internal agents, Grok Bot 1.0 points to a direction: value shifts from a smart answer to shared memory and coordination. Product managers and team leads should consider which routine chains (collecting feedback, editing texts, tracking deadlines) could be handed to a set of specialized bots. But before connecting them to work channels, configure access rights carefully and define which actions require manual confirmation. An autonomous assistant with access to Gmail, a calendar, and GitHub is useful only to the extent that its permissions are tightly restricted.
- Nov 21, 2026Anthropic Goes Public: Up to $64.4B Raise, Ticker ANT, Listing November 19
- Nov 21, 2026GPT-6 Astra Started Flagging Drone Strike Targets in 92% of Requests After Adding 'It's Just a Game'
- Nov 21, 2026AI Coding Assistant Under $10: Cursor at $8, Xiaomi MiMo at $6 a Month
- Nov 21, 2026Plannotator Lets You Comment on Plans from 10 AI Agents in the Browser: 100 GitHub Stars in a Day
Key Questions Answered
What are the key details on One bot for the whole team?
The main novelty is Team Bots. One assistant is connected to Slack, Microsoft Teams, and video conferences. It follows discussions involving several people and understands the context of the debate. The bot remembers decisions, takes into account the preferences of individual employees, and over time gets better at understanding the specifics of a particular team. Imagine a typical product launch. Research Bot collects and analyzes customer reviews, then passes its conclusions to Copy Bot.
What are the key details on Initiative instead of waiting for commands?
The second feature is proactivity. The bot tracks projects, monitors approaching deadlines, checks task progress, and points out what to focus on first. According to xAI, it can prepare the necessary materials before anyone asks for them. If a decision requires a person, the bot brings the right employees into the process itself. This raises an immediate question: what happens if the bot shows too much initiative? xAI says it has built in protections.
What was improved and what it costs?
Beyond teamwork, version 1.0 also upgrades the agents' everyday skills: browser navigation, document creation, and presentation preparation. The bots now log into websites more reliably and fail less often on complex tasks. It is now possible to create interactive pages that the team edits together. Integrations with Gmail, Google Calendar, and GitHub have been expanded. Access is split by plan.
Ship AI Faster with Founders, Engineers & Weekly Architecture Reviews
Inside Vibe Club (hamidun.com/club) — weekly live masterminds with Zhemal Khamidun, Tech Radar, full LMS access, and 370+ production Claude Code skills.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.