Anthropic launched Claude Meetings on 10 November 2026. Claude joins a video call as a visible participant. While people talk, it can draft the plan under discussion, map a timeline on a whiteboard, answer questions in the meeting chat, and retrieve information from connected work apps, such as last year’s figures.

It is in beta on Team and Enterprise plans, and at launch it works with Google Meet. After the call, the notes, action items and anything else Claude made are available in claude.ai. Anthropic says teams can “turn discussions into actions in real time.”

The excitement is about the live demo

A model drafting a plan while you talk looks impressive, and it is. A timeline appearing on a whiteboard mid-call makes for a good screen recording.

I separate “technically impressive” from “where this is actually going”. Both can be true of the same product. So I ask the first question I ask of any launch. Is there a need for it? What problem does this solve that I couldn’t already solve?

Meetings have a real problem. Someone agrees to do something on a call, nobody writes it down properly, and a week later the work hasn’t started. A tool that captures the plan as people speak addresses that.

What changed and what didn’t

What changed is that the drafting and note-taking now happen inside the call, and Claude can pull figures from your connected apps without someone leaving to look them up.

The article says Claude produces action items that sit in claude.ai, and that it can retrieve information from connected apps. It never says Claude creates tasks in your other apps. Reading from your systems is one thing. Writing into them is another.

That gap matters. A meeting pays off only when the plan becomes tasks that run in your systems, with owners and dates, in the tools your team already checks every morning. If the action items stay in a separate place, you have a tidier summary. A tidy summary is better than a messy one, but it still needs a person to carry it into the work.

So I would judge Claude Meetings by what got done after the call, and ignore the whiteboard.

One assistant used well beats a separate meeting tool

Take a service business with a small team. A clinic with four staff, say, planning a new booking process for the next quarter. They hold a weekly meeting. Someone has to turn what was said into things that happen.

The question for them is whether adding another tool to the stack is worth it, because a good meeting assistant alone doesn’t settle that. I find that one tool used deeply is more efficient and more effective than a pile of them. I use Claude for almost everything, so I have a bias here, and I will say so. But the reasoning holds either way. Every extra tool brings another login, another place where information lives, and another habit to build.

If your team already works with Claude, then a meeting feature that shares the same context and the same connected apps has a real advantage over a standalone meeting tool. The plan drafted in the call lives where the rest of your work with Claude lives. If your team does not use Claude yet, the case is weaker. You would be adopting a new assistant to get a meeting feature, and that is a bigger decision than it looks.

Also note the limits in the launch. It is on Team and Enterprise plans only. It works with Google Meet at launch. Claude won’t join calls with external guests unless an admin authorises it, and it leaves automatically if an external guest joins. A clinic that meets with suppliers or outside consultants will hit that boundary quickly.

Where this connects to what I build

This is the same question I ask about customer conversations. A chat that answers a question is useful. A chat that checks the order, changes the booking or escalates to your team, inside your systems, is where the work gets done. That is the idea behind AeroChat, which connects to your own data systems and backend processes so it can perform actions, not just answer questions. Meetings and customer chats are different settings, but the test is the same. Did the conversation end in something that happened?

What a business should do

Don’t judge the product by the demo. Run it on one recurring meeting for a few weeks, and only if your team already works in Claude.

Before the trial, write down what you want out of the call. For example, five tasks, each with an owner and a date. After the call, count how many of those tasks exist in the system where your team actually tracks work, and how many someone had to retype or copy over by hand.

If most of them landed without manual effort, keep it. If people are copying action items out of claude.ai into your task tracker, you have bought a summary tool. In that case, stay with the one assistant you use well and a short follow-up message after each call.

Source: TechCrunch: Anthropic launches Claude Meetings to turn workplace discussions into plans and tasks