How to Remove an AI Notetaker From a Meeting You Are In

How-to — 2026-11-14 — by Mahmoud Zalt

Get a recording bot out of a live call, including one you did not invite. What works per platform, what is already captured, and how to ask well.

You join a call and there is a participant that is not a person. Nobody mentioned it, the conversation is about to get candid, and you have roughly nine seconds to decide whether saying something is worse than not saying something.

The mechanics are easy. What follows is the removal steps per platform, what has already been captured by the time you act, what can and cannot be blocked in advance, and how to raise it without making the call weird.

Are you allowed to remove a notetaker you did not invite?

Technically, if you have the permission, yes. A notetaker joins as an ordinary participant, so whatever rights let you remove a person let you remove the bot. Otter goes further and documents that any meeting participant, including those without an Otter account, can remove its Notetaker.

Socially, it depends on how you do it. Ejecting someone's bot without a word reads badly, especially if they are a client. Saying you would rather keep this one off the record, before the substance starts, almost always lands fine and takes one sentence.

It also helps to remember why the bot is often there. Plenty of people send a notetaker because they could not attend, not because they are gathering evidence. Asking first separates those two cases, and it usually turns out to be the first one.

How do you remove it on each platform?

Open the participant list and treat the bot as a guest. Otter and Fireflies both document this path across Zoom, Google Meet and Microsoft Teams, and in every case you start from the participant list and use the menu next to the notetaker's name.

Any Otter Notetaker: type stop otter in the meeting chat. Otter documents that this removes every Notetaker in the meeting at once and needs no Otter account.
Zoom: the host or co-host opens the participant list, uses the menu next to the bot's name, and removes it.
Google Meet: depending on the meeting permissions, hosts and other participants can remove it from the call via the participant list.
Microsoft Teams: depending on the meeting permissions, the organiser and presenters can remove it from the meeting.
From the owner's dashboard: if the account is yours, both tools document a stop control that ends the capture without touching the participant list.

One caveat on the chat command. Because stop otter removes every Otter Notetaker present, it is a blunt instrument on a call where several people brought one. If you only want one gone, use the participant list instead.

Getting the bot out without wrecking the call

Name it early, before the real conversation starts — One line does it. Say you noticed a notetaker and ask whether it can sit this one out. Doing it in the first minute is normal. Doing it twenty minutes in, after something sensitive, is what makes it awkward.
Ask the person who brought it to remove it — They can stop it from their own dashboard in a couple of clicks, and it keeps the decision with the person whose account holds the recording. This is the first move whenever you are not the host.
Use the chat command if it is an Otter bot — Typing stop otter in the Zoom, Google Meet or Microsoft Teams chat removes every Otter Notetaker immediately. Otter documents that no account is needed, so this works even when you are a guest on someone else's call.
Remove it from the participant list if you have the rights — Host or co-host on Zoom, and the equivalent permission on Google Meet or Teams. Treat it exactly like removing any other participant, using the menu beside the bot's name.
Take the address off the calendar invite for next time — If a notetaker address is sitting on a recurring event's guest list, it keeps rejoining regardless of anything said in a single call. Removing it from the invite stops the next occurrence.
Agree who takes notes instead — The reason the bot was there does not disappear when it leaves. Offer to send a short summary afterwards and the whole exchange stops feeling like a refusal.

What was already recorded by the time you removed it?

Some of it, depending on how fast you were and which tool it is. Fireflies documents that it does not process meetings under three minutes, so a removal inside that window produces no transcript and no notes at all. After that, it delivers notes covering the portion it recorded.

Whatever was captured lives in the account that owns the bot, not in yours. That means you cannot delete it, and asking the account holder is the only route. Most people say yes to that request without much thought, particularly if you ask in the moment.

This is the practical argument for speaking up in the first minute rather than the twentieth. It is not really about etiquette. It is that a fast removal leaves nothing to negotiate over afterwards.

Can you block notetakers before they turn up?

Partly, and it is worth being straight about the limits. The waiting room is the only control that reliably catches every unknown participant, because it stops anyone who is not on your list regardless of what they are. Everything else covers some cases and misses others.

On the documented side, Otter Enterprise supports blocked keywords that stop its Notetaker auto-joining meetings whose titles match, applied across a whole workspace. Otter states this is not yet available in their interface, so an admin has to request it through support, and that it only covers members of your own workspace.

The rest is community-reported rather than vendor-documented. In a long public Zoom Community thread, an admin who had already blocked notetaker domains and required authenticated participants still had bots joining every meeting. Another member relayed guidance from Zoom support pointing to the Zoom App Marketplace, under Admin App Management, where an admin can manage or remove a third party app for the account.

A later reply in the same thread describes turning off app requests in Admin App Management, which then surfaced a prompt allowing the admin to decline the request and decline future ones. On the Microsoft side, a Microsoft Q&A thread answered by a Microsoft moderator points admins to the Entra admin centre, where an enterprise app can be set so users cannot sign in to it.

Where does consent actually sit?

With the person doing the recording. Rules differ by country and by state, some requiring every participant to agree and others requiring only one, and the obligation belongs to the recorder rather than to the software. On calls that cross borders, the safe assumption is the strictest rule in the room.

That cuts both ways, which is the useful part. If you brought the bot, announce it. If someone else did, you are entitled to ask what is being recorded and where it goes, and a reasonable counterpart will just tell you.

What was the notetaker there to do anyway?

Usually not to keep a record. It was there because somebody needs the follow-up sent, the commitment tracked and the next step booked, and a transcript was the closest thing available. Which is why removing it solves the privacy problem and none of the work problem.

That is what we went after at Sistava. Instead of a bot sitting in the call, you hire an AI Employee, tell it in plain English what should happen after your meetings, and it works your calendar and your inbox from your own side of the conversation.

What removal will not achieve

It will not stop someone recording if they intend to. As several people in that Zoom thread pointed out, a phone on the desk does the same job, and no meeting platform setting reaches it. Removing a bot changes what is convenient, not what is possible.

It will not delete what was already captured either, since that sits in someone else's account. And it will not answer whether the recording was reasonable. Plenty of notetakers are there for entirely ordinary reasons and their owner would have removed it happily if asked.

Which points at the thing that actually works. A short norm at the start of a call, saying whether anything is being recorded and why, prevents more of this than any admin setting. Tools follow the agreement, not the other way round.

FAQ

Can I remove an AI notetaker if I am not the host?

Sometimes. If it is an Otter Notetaker, Otter documents that any participant can type stop otter in the meeting chat to remove it, with no Otter account required. Otherwise you need the platform permission to remove participants, which on Zoom means host or co-host. If you do not have it, ask the person who brought the bot, or ask the host to remove it.

How do I remove a notetaker from Zoom, Google Meet or Microsoft Teams?

Open the participant list and use the menu next to the bot's name, exactly as you would for a person. On Zoom the host or co-host can do this. On Google Meet and Microsoft Teams it depends on the meeting permissions, with organisers and presenters typically able to remove participants on Teams.

Was the meeting already recorded before I removed the bot?

Partly, depending on timing and tool. Fireflies documents that meetings under three minutes are not processed, so a fast removal produces no transcript at all, and that a later removal still yields notes for the portion recorded. Anything captured lives in the account that owns the bot, so only that account holder can delete it.

Can an admin block AI notetakers from all our meetings?

Only partly. A waiting room is the one control that catches every unknown participant. Otter Enterprise documents keyword blocking by meeting title, configured through Otter support, that applies to your workspace only. Community threads report that domain blocks and authentication requirements help but do not catch everything, since some notetakers join through a participant's own authenticated account.

Is it rude to remove someone's notetaker?

Not if you say something. Removing a bot silently reads badly, but asking in the first minute whether the call can stay off the record is a normal request that most people accept without friction. It also helps to offer a short written summary afterwards, since that is usually what the notetaker was there to produce.

Do I have to be told when a meeting is being recorded?

It depends on where you and the other participants are. Some jurisdictions require every participant to consent, others require only one, and the obligation sits with whoever is recording rather than with the software. Notetaker bots normally appear in the participant list under their own name, which is a signal but not a substitute for someone saying it out loud.