AI tools don’t fail because the technology is bad. They fail because nobody remembers to use them.
That extra tab, a separate login, the habit that never quite sticks. That’s where most AI investment quietly disappears.
It’s a pattern I see regularly. The intention is there. The licence is paid for. But six months in, a handful of people are using it well and everyone else has drifted back to doing things the old way.
That’s not a people problem. It’s a design problem. AI that lives outside the flow of work will always struggle to become part of it.
Anthropic’s new Claude Tag, launched on 23 June 2026, is designed to fix that problem at the source. Rather than living in a separate app, Claude joins your Slack channels as a team member, builds context over time, and gets to work without being prompted every time.
It reads what’s happening in your channels, picks up on tasks, and responds when it’s useful, either when someone asks or when it spots something worth flagging.
That might sound like a subtle change. But when you think about how much of your team’s actual work happens inside a chat tool, the idea of AI that’s already there, already paying attention, and already ready to help is a meaningful shift in how AI fits into the working day.
It’s a different way of thinking about what AI should do at work, and it’s worth understanding even if Slack isn’t part of your current setup.
What Claude Tag Does
Claude Tag lets you invite Anthropic’s Claude AI into a Slack channel the same way you’d add a colleague. Once it’s in, anyone on the team can type @Claude and delegate a task, ask a question, or request an update. Claude reads the conversation in that channel, builds context over time, and responds inside a thread, including working through multi-step tasks using whatever tools and data it’s been given access to.
There are two modes. The first is on-demand, where someone tags Claude directly and gives it a job. The second is “ambient mode”, where Claude can proactively jump into a conversation when it spots something worth flagging: a missed follow-up, a relevant update from elsewhere in the business, or a task that’s gone quiet.
It’s currently available in research preview for Claude Enterprise and Team customers, and Anthropic has said it will expand to other platforms beyond Slack in the future.
A few things make this different from AI you’ve probably already tried:
Shared context across a team. Every person in the channel works with the same Claude identity, so one person can pick up exactly where another left off. No repeating instructions, no context loss.
Scoped access set by admins. Business leaders and IT teams define which channels and data Claude can see. A Claude set up for your legal team can’t read your engineering channel, so the boundaries are clear and configurable.
It learns your business over time. Claude builds a persistent memory of the context in the channels it’s part of, meaning the longer it works with your team, the more relevant its responses become.
This Isn’t Just About Slack
If you don’t use Slack, you might be tempted to scroll past this. I’d encourage you not to.
Claude Tag is an early, visible example of a category of AI that’s been talked about for the past year or two but is now arriving in real products: the AI teammate.
Not a tool you prompt. Not a chatbot embedded in an app. An AI that lives inside the workflow, monitors what’s happening, takes on tasks, and follows up the way a capable person on your team would.
Microsoft has been building toward this for a while. Copilot Cowork, which went generally available on 16 June 2026, is Microsoft’s own answer to the same idea: an agent-style experience that works across Teams and Microsoft 365 to coordinate tasks, track context, and act proactively.
What Anthropic is doing with Claude Tag is proving that this model of AI works in an enterprise messaging environment. That matters for every business leader thinking about where AI fits into day-to-day team collaboration, regardless of which tools they currently use.
Three Questions Worth Asking Before You Enable It
This isn’t a warning against trying AI teammates. I think they’re worth exploring seriously. But like any tool that has access to your team’s conversations and data, it’s worth going in with a clear head.
What data is it going to see?
Claude Tag’s value comes from context, and the more channels it’s in, the more useful it becomes. But that same breadth of access means you need to be deliberate about governance before you start.
What channels contain sensitive information? What client data, personnel matters, or commercial discussions happen in Slack that you wouldn’t want an AI reading indiscriminately? Defining access boundaries up front is essential.
The introduction of ambient mode, where Claude can jump in without being asked, is interesting, but it also raises real change management questions. Some people will find proactive AI input helpful.
Others might find it jarring, or worry about what it says about the messages they’re sending.
Having a clear, communicated policy for how AI is used in your team’s chat tools matters more than the technology itself.
Are you ready to think about Al teammates across platforms?
If you’re already running Microsoft 365, you likely have Copilot-capable tools available to you right now. The smarter move for most Australian businesses isn’t to add a second AI ecosystem.
It’s to get more from the one they have, with proper configuration, security controls, and adoption support in place. That said, understanding what Anthropic is doing here helps you ask better questions of Microsoft too.
What This Means for Australian Business Leaders
The pace at which AI teammates are arriving in collaboration tools is faster than most businesses have planned for. Claude Tag isn’t isolated. It’s one signal in a pattern that includes Microsoft Scout on the desktop, Copilot Cowork in Teams, and a growing set of agents that can schedule work, follow threads, and take action without waiting to be prompted.
For business leaders, this means three practical things:
Your collaboration tools are about to become AI environments. Whether it’s Slack, Teams, or something else, the question isn’t whether AI will be present in your team’s conversations. It’s on what terms, with what access, and with what guardrails.
Governance and data structure matter more than ever. We’ve covered this before in the context of Microsoft 365: if your data isn’t in good shape, your AI results won’t be either. The same principle applies here. Claude Tag learns from the content in your channels, so what’s in those channels and how it’s organised shapes what the AI can do.
Change management is the harder job. The technology is mostly ready. The human side, helping your team understand what AI teammates do, building trust in how they’re used, and setting clear expectations, is where most businesses under-invest. Doing that well takes time and a clear plan.
Where to From Here
Claude Tag is in research preview right now. It’s not something most Australian businesses need to act on today, but it is something worth understanding because the pattern it represents is moving fast.
If you’re still building the foundation, working out how to deploy Copilot well, how to manage your Microsoft 365 data, or how to help your team actually adopt the AI tools you’ve already paid for, that’s the right place to put your energy.
The more that foundation is in order, the better placed you’ll be when AI teammates become a standard part of the modern workplace.
If you’d like to talk through what this looks like for your business, or how CG TECH can help you build an AI-ready environment across Microsoft 365 and beyond, get in touch with the team.
About the Author
Carlos Garcia is the Founder and Managing Director of CG TECH, where he leads enterprise digital transformation projects across Australia.
With deep experience in business process automation, Microsoft 365, and AI-powered workplace solutions, Carlos has helped businesses in government, healthcare, and enterprise sectors streamline workflows and improve efficiency.
He holds Microsoft certifications in Power Platform and Azure and regularly shares practical guidance on Copilot readiness, data strategy, and AI adoption.
AI tools don’t fail because the technology is bad. They fail because nobody remembers to use them.
That extra tab, a separate login, the habit that never quite sticks. That’s where most AI investment quietly disappears.
It’s a pattern I see regularly. The intention is there. The licence is paid for. But six months in, a handful of people are using it well and everyone else has drifted back to doing things the old way.
That’s not a people problem. It’s a design problem. AI that lives outside the flow of work will always struggle to become part of it.
Anthropic’s new Claude Tag, launched on 23 June 2026, is designed to fix that problem at the source. Rather than living in a separate app, Claude joins your Slack channels as a team member, builds context over time, and gets to work without being prompted every time.
It reads what’s happening in your channels, picks up on tasks, and responds when it’s useful, either when someone asks or when it spots something worth flagging.
That might sound like a subtle change. But when you think about how much of your team’s actual work happens inside a chat tool, the idea of AI that’s already there, already paying attention, and already ready to help is a meaningful shift in how AI fits into the working day.
It’s a different way of thinking about what AI should do at work, and it’s worth understanding even if Slack isn’t part of your current setup.
What Claude Tag Does
Claude Tag lets you invite Anthropic’s Claude AI into a Slack channel the same way you’d add a colleague. Once it’s in, anyone on the team can type @Claude and delegate a task, ask a question, or request an update. Claude reads the conversation in that channel, builds context over time, and responds inside a thread, including working through multi-step tasks using whatever tools and data it’s been given access to.
There are two modes. The first is on-demand, where someone tags Claude directly and gives it a job. The second is “ambient mode”, where Claude can proactively jump into a conversation when it spots something worth flagging: a missed follow-up, a relevant update from elsewhere in the business, or a task that’s gone quiet.
It’s currently available in research preview for Claude Enterprise and Team customers, and Anthropic has said it will expand to other platforms beyond Slack in the future.
A few things make this different from AI you’ve probably already tried:
This Isn’t Just About Slack
If you don’t use Slack, you might be tempted to scroll past this. I’d encourage you not to.
Claude Tag is an early, visible example of a category of AI that’s been talked about for the past year or two but is now arriving in real products: the AI teammate.
Not a tool you prompt. Not a chatbot embedded in an app. An AI that lives inside the workflow, monitors what’s happening, takes on tasks, and follows up the way a capable person on your team would.
Microsoft has been building toward this for a while. Copilot Cowork, which went generally available on 16 June 2026, is Microsoft’s own answer to the same idea: an agent-style experience that works across Teams and Microsoft 365 to coordinate tasks, track context, and act proactively.
If you’ve been following the rise of agentic AI systems in 2026, you’ll recognise that Claude Tag is part of the same wave.
What Anthropic is doing with Claude Tag is proving that this model of AI works in an enterprise messaging environment. That matters for every business leader thinking about where AI fits into day-to-day team collaboration, regardless of which tools they currently use.
Three Questions Worth Asking Before You Enable It
This isn’t a warning against trying AI teammates. I think they’re worth exploring seriously. But like any tool that has access to your team’s conversations and data, it’s worth going in with a clear head.
What data is it going to see?
Claude Tag’s value comes from context, and the more channels it’s in, the more useful it becomes. But that same breadth of access means you need to be deliberate about governance before you start.
What channels contain sensitive information? What client data, personnel matters, or commercial discussions happen in Slack that you wouldn’t want an AI reading indiscriminately? Defining access boundaries up front is essential.
We’ve written before about why AI data governance deserves more attention than it usually gets, and Claude Tag is a good prompt to revisit that conversation for your own environment.
How will your team actually use it?
The introduction of ambient mode, where Claude can jump in without being asked, is interesting, but it also raises real change management questions. Some people will find proactive AI input helpful.
Others might find it jarring, or worry about what it says about the messages they’re sending.
Having a clear, communicated policy for how AI is used in your team’s chat tools matters more than the technology itself.
Are you ready to think about Al teammates across platforms?
If you’re already running Microsoft 365, you likely have Copilot-capable tools available to you right now. The smarter move for most Australian businesses isn’t to add a second AI ecosystem.
It’s to get more from the one they have, with proper configuration, security controls, and adoption support in place. That said, understanding what Anthropic is doing here helps you ask better questions of Microsoft too.
What This Means for Australian Business Leaders
The pace at which AI teammates are arriving in collaboration tools is faster than most businesses have planned for. Claude Tag isn’t isolated. It’s one signal in a pattern that includes Microsoft Scout on the desktop, Copilot Cowork in Teams, and a growing set of agents that can schedule work, follow threads, and take action without waiting to be prompted.
For business leaders, this means three practical things:
Where to From Here
Claude Tag is in research preview right now. It’s not something most Australian businesses need to act on today, but it is something worth understanding because the pattern it represents is moving fast.
If you’re still building the foundation, working out how to deploy Copilot well, how to manage your Microsoft 365 data, or how to help your team actually adopt the AI tools you’ve already paid for, that’s the right place to put your energy.
The more that foundation is in order, the better placed you’ll be when AI teammates become a standard part of the modern workplace.
If you’d like to talk through what this looks like for your business, or how CG TECH can help you build an AI-ready environment across Microsoft 365 and beyond, get in touch with the team.
About the Author
Carlos Garcia is the Founder and Managing Director of CG TECH, where he leads enterprise digital transformation projects across Australia.
With deep experience in business process automation, Microsoft 365, and AI-powered workplace solutions, Carlos has helped businesses in government, healthcare, and enterprise sectors streamline workflows and improve efficiency.
He holds Microsoft certifications in Power Platform and Azure and regularly shares practical guidance on Copilot readiness, data strategy, and AI adoption.
Sources
Recent Posts
Popular Categories
Archives