Coworkers gather around a friendly AI robot displaying productivity tools and cybersecurity icons in a modern office.
SpaceXAI wants your next hire to be a bot, and it wants the whole department to share it. On Monday, September 28, 2026, the company launched Team Bots. A Team Bot is a Grok Bot that a whole team shares. Each one is built around a job or a workflow, keeps what it learns, and connects to business apps with plugins and credentials. For IT admins and developers, this is another AI agent that can reach company systems, so it needs the same review you'd give any new user with a password.

What SpaceXAI actually launched​

The company's announcement calls Team Bots Grok Bots that work and learn alongside your team. The idea is simple: give one access to the files, apps, and expertise it needs, then share it so everyone can work from the same context.

Most AI assistants are one person having one chat. A Team Bot is set up for a role instead. SpaceXAI says you build a Team Bot around a role or workflow your team shares, and everyone gets access to the same Bot and its expertise, while still being able to work with it individually.

The basics:

  • Status: Public beta
  • Who can use it: Teams and Enterprise plans
  • Starter bots: Pre-built versions for sales, product management, marketing and data analytics
  • Where you use it: In Slack or in the Grok Bot app
  • Not announced: A release version, separate Team Bots pricing, regional availability, minimum seat counts or a general availability date

TeslaNorth reported the launch the same day and confirmed that Team Bots are available now in public beta for Teams and Enterprise plans, with pre-built versions available for sales, product management, marketing and data analytics.

Summary: This is a shared, role-based AI agent in public beta for business plans. It isn't a consumer chatbot feature.

How a Team Bot is built​

SpaceXAI says every Team Bot has four parts. Anyone who manages access to business systems should read this list closely:

  1. Context: Files, instructions and "skills," such as brand guides or internal documentation.
  2. Plugins: Connections to apps such as Salesforce, Notion and GitHub. SpaceXAI says each person can connect a plugin, or it can be set up for the whole team.
  3. Credentials: API access to third-party services that don't have a plugin. The company says the bot accesses them securely.
  4. Memories: What the bot learns on the job, which the company says makes it better at its role over time.

In practice, context and memories are what the bot knows. Plugins and credentials are what it can reach and change. That second half is what administrators need to watch.

The privacy line SpaceXAI is drawing​

The company says sharing a bot doesn't mean sharing your chats. Each person's conversations with a shared bot stay private. The bot keeps separate context and memories for each user while drawing on the skills shared across the team.

That's a claim from the company, not an audited guarantee. The announcement doesn't say:

  • What administrators can see or inspect
  • How team-shared memories are created, reviewed or corrected
  • How long data is kept
  • Whether the privacy behavior differs by plan
  • How credentials are stored, or what audit logs exist

Runtimewire, another outlet covering the launch, described Team Bots as coming with shared skills, private user conversations and administrator controls. SpaceXAI's own announcement doesn't spell out what those admin controls are. Until documentation appears, treat "administrator controls" as something to check during a trial.

There's also a catch in SpaceXAI's own examples. Its Data Bot "remembers corrections" so that one person's feedback improves answers for everyone. So some learning clearly does cross between users. How that fits with separate per-user memories isn't explained. It's one of the first questions to ask a sales rep.

Summary: SpaceXAI says your chats stay private and team skills are shared. The details that matter most to compliance teams aren't documented yet.

Slack gets it now. What about Microsoft Teams?​

Team Bots launched with Slack support. SpaceXAI says each bot has its own handle, so you can invite it to a channel where members ask questions, add context and see its answers.

Microsoft shops should be careful with the details. Grok Bot already had a Microsoft Teams connector before this launch. TeslaNorth reported on September 11, 2026 that Grok Bot could search, read and send Microsoft Teams chats and channel messages. SpaceXAI's official plugin listing describes the same abilities.

That connector is an older, separate Grok Bot feature. SpaceXAI announced Team Bots for Slack and the Grok Bot app. The launch materials don't say Team Bots can be invited to Teams channels the way they can join Slack channels. They also don't say whether the Teams connector uses the same permissions or privacy model. And nothing suggests Microsoft built, endorses or runs any of it.

In short, Slack gets the bot as a channel member at launch. Teams users get a connector the bot can use. Those are different things.

How SpaceXAI says it uses Team Bots​

The company gave four examples from its own teams. All the results below are reported by SpaceXAI or its customers and haven't been independently checked.

Sales and customer success​

Each major sales account has its own bot. The account executive, customer success manager, solutions architect and sales leader all share it. Overnight, the bot reads company news, recent Gong calls, Notion docs and relevant Slack threads. Each morning it posts a briefing in the account's Slack channel with drafts written for each role. SpaceXAI says the bot becomes the account's system of record as people rotate on and off.

SpaceXAI also quotes a customer, insurer Harper. Its CEO, Dakotah Rice, says the company built a bot in 24 hours to replace manual checks across three platforms. He says it saved customers more than $120,000 across hundreds of lapsed policies. Those figures come from Harper as quoted by SpaceXAI.

Product and engineering​

Developers will want to read this one. The engineering bot works from a project's Slack channel and connects to Notion, Linear, Hex, Datadog and Cursor. It follows product decisions, triages bug reports, creates tickets and launches Cursor Cloud Agents to handle well-defined fixes.

SpaceXAI says it taught the bot its shipping process through skills: how to handle PR reviews, when to ask the team for help, and what evidence a change needs before it's done. The company says the bot steered a Cursor project that ran hundreds of cloud agents. That helped a five-person team ship more than 100 pull requests a day and launch Team Bots "in a few weeks."

That's a striking number, but it's the company describing its own product, built by its own team, with its own tools. Pull-request counts also say nothing about code quality, review load or how many fixes had to be rolled back. It's a marketing claim, not a benchmark.

Marketing​

The marketing bot checks drafts posted in Slack against brand guidelines, earlier blog posts and social copy. For website and SEO updates, it makes the approved change itself and posts a preview link for final sign-off. So this bot doesn't just give advice. It can edit a public website, with a human approving the last step.

Data analytics​

The Data Bot uses shared, read-only credentials to query approved Databricks tables. It also connects to Datadog, Hex and Statsig. SpaceXAI says its analytics team spent two years building a library of skills covering more than 45,000 tables and handed it to the bot. The "read-only" part applies to the warehouse credentials only. The announcement doesn't say every connected system is read-only. In fact, one of the listed skills covers safely updating company-wide dashboards.

Summary: The examples range from read-only lookups to bots that file tickets, launch coding agents and change live websites. That range is exactly why governance matters here.

Where Team Bots fit in SpaceXAI's lineup​

Team Bots build on Grok Bot, SpaceXAI's agent platform. Its own announcement said Grok Bot entered beta in August for various SuperGrok and Cursor subscribers on desktop and iOS, with enterprise users placed on a waitlist. Engadget describes Grok Bot as a digital work assistant that signs into tools like Gmail, LinkedIn and Salesforce and acts on your behalf. It also notes that Grok Bot was originally the brain child of Cursor, codenamed Sand, and partially built on SpaceXAI's compute.

For Windows users: VentureBeat reported that Grok Bot arrived for macOS, Windows, Linux and iOS, with Android listed as coming soon. Engadget's more recent guide says it can be downloaded on macOS, Windows, iOS and Android devices. Neither report says anything specific about how Team Bots behave on Windows.

Pricing is unclear, and the reports don't agree. VentureBeat reported in August that Cursor Premium Teams costs $120 per seat per month and adds centralized billing and settings, a team marketplace for skills and plugins, shared usage analytics and SAML/OIDC single sign-on. Runtimewire says the Grok Bot page lists Cursor Teams at $40 per seat per month, and that the launch announcement does not describe a separate Team Bots charge. Plan names and prices seem to be changing, so get a written quote before budgeting.

What IT admins should check before switching it on​

This part is my own advice as an IT journalist, based on how the product is described. None of these are controls SpaceXAI has confirmed.

  1. Check every plugin's permissions. A Salesforce or GitHub plugin set up for the whole team carries whatever permissions it was granted. Find out whether team plugins run under a shared service identity or each user's own account.
  2. Treat credentials like service accounts. Give them the least access possible, rotate them, and keep a record of who owns them. SpaceXAI's Data Bot is a good model here: read-only, approved tables only.
  3. Map anything that can make changes. List every bot that can send messages, file tickets, launch coding agents or edit websites. Each one needs a named human who approves its work.
  4. Ask who controls shared memory. If one person's "correction" teaches the bot something wrong, how do you find it and undo it?
  5. Get privacy answers in writing. "Conversations remain private" should come with clear answers on retention, admin access and eDiscovery before any regulated data goes near it.
  6. Watch for shadow IT. Engadget recommends testing a few low-risk tasks at first, and keep human approval enabled for any actions that could potentially fire off messages. That advice matters even more when a bot serves a whole team.

The bigger picture​

Shared AI agents that live in team chat, hold company knowledge and act in business apps are the category Microsoft is chasing with Copilot and its own agent tools. It's a crowded race. SpaceXAI's pitch is that one bot, loaded with a team's process and memory, can outlast staff turnover and keep a project's history in one place.

Whether that's an advantage or a risk depends on governance, and the beta leaves most of those questions open. A bot that remembers everything, reaches into Salesforce and GitHub, and can ship code while you sleep can be very productive. It also needs someone watching the logs.

As TeslaNorth put it, the next test is whether companies treat these bots as occasional assistants or begin assigning them persistent roles in everyday work. My advice for IT departments: keep the trial small, give plugins and credentials the minimum access they need, and get the privacy promises in writing.

 

References

  1. SpaceXAI Launches Team Bots That Learn Alongside Your Workforce - TeslaNorth.com TeslaNorth.com 2026-09-28T22:00:33+00:00
  2. SpaceXAI opens Team Bots beta for shared AI work across companies runtimewire.com
  3. SpaceXAI's Grok Bot turns agents into persistent digital coworkers that can operate your apps for $120-per-month | VentureBeat venturebeat.com