AI Exposure · Meeting notetaker
What Zoom AI Companion actually does with your data
The AI inside the video calls much of the sector runs on. Zoom's no training promise is unusually specific, and there is a reason it is worded that way.
- Trains on your data
- Zoom says no, in writing
- Processed
- Zoom plus third party models
- Can you turn it off
- Yes, and off by default
- Verified
- 2 July 2026, against Zoom's published documents
What AI is in it?
- Meeting summaries. Recaps the meeting and sends a summary afterwards, so a record of what was said exists even if nobody pressed record.
- In-meeting AI questions. Anyone who arrives late can ask the AI what they missed, from a side panel, using the live conversation as its source.
- Drafting help across Zoom. Composes chat replies and tidies text inside the wider Zoom Workplace apps.
Does your data train AI models?
Zoom says no, in writing.
Zoom's published position is specific: "Zoom does not use any customer audio, video, chat, screen-sharing, attachments, or other communications-like customer content to train Zoom's or third-party artificial intelligence models."
There is history behind that wording. In 2023 Zoom's terms of service appeared to permit the opposite, customers noticed, and the current commitment was written in response.
Two things follow. The promise is unusually precise because it had to be. And the terms changed quietly enough that it took readers, not an announcement, to catch it.
Where is your data processed?
Zoom plus third party models.
Zoom uses what it calls a federated approach: its own AI models plus third party ones, naming OpenAI, Anthropic and Meta's Llama. So what someone says in your meeting can be processed by companies other than Zoom, under Zoom's agreements with them.
Zoom does not publish one tidy page saying where every AI feature is processed; the realistic assumption for a UK charity is the United States unless your admin has set up one of Zoom's data residency options.
What changes between free and paid?
Paid plans, off by default.
AI Companion is included at no extra cost on paid Zoom plans and is not part of the free tier.
The features arrive switched off: an administrator has to enable them at account level before anyone sees a summary. That is the opposite of the usual default, where the AI appears and someone has to notice it.
Can your organisation turn it off?
Yes, and off by default.
Account owners and admins enable or disable each feature, at account and group level, and everything starts disabled. Hosts control it in the meeting, and Zoom says participants get visibility when AI features are active.
So the question is not whether your organisation can control it. It is whether anyone has looked at the settings since the account was set up.
What does this mean for the people you support?
An AI summary of a supervision or safeguarding call is a written record that did not exist before, about people who may not know it exists. The settings here are yours to set, which moves the whole question out of the vendor's hands and into your meeting rules.
An off switch does not decide which meetings AI is allowed to attend. Somebody at your organisation has to.
What to check this week
- Ask what is actually enabled. Whoever owns your Zoom account should check Account Settings, then AI Companion, and tell you which features are on and for whom. Off by default only helps if nobody has enabled it casually since.
- Set the meeting rule. Name the meeting types where no AI summary is ever welcome: safeguarding, casework, supervision, HR. The rule belongs in your AI policy, and it is the same rule as for Otter and Teams.
- Ask the third party question before enabling. If it matters to you which companies process your meetings, ask Zoom which AI Companion features send content to which model providers, and whether a data residency option covers your plan.
Sources
Everything above is my plain English reading of what Zoom publishes. The originals:
- Zoom: How Zoom AI features handle your data reviewed 2 July 2026
- Zoom: AI Companion announcement (no additional cost, federated models) reviewed 2 July 2026
Vendors change these documents quietly and often. If you spot something out of date, email me and I'll re-verify the entry.
Who invited it to the meeting?
The settings above are worth an afternoon. But the real exposure with a notetaker is not a switch. It is which meetings it sits in, because meetings are where the things too sensitive to write down get said out loud.
And several answers above split on whose account it joins from. Anyone in a meeting can bring a notetaker, including guests from outside your organisation, on free personal plans nobody at your end can see.
So this is a rule to agree, not a box to untick: which meetings can be recorded, who has to say yes first, and whose account the recording lives in.
My free Charity AI Policy Generator writes that rule with you in a few minutes. And finding out what has already been sitting in on your meetings is what a discovery is for.
See your whole toolkit at once
This page covers one tool. The AI Exposure Report Generator covers your toolkit: tick the tools your charity uses and get one report showing what the AI inside each of them does with your data, as a Word document you can table at your next meeting. Free, and nothing you tick leaves your browser.
The Report Generator, start to finish: tick your tools, read your exposure, download the report.
Read the other tools
The same questions, answered for every tool in the directory.
Your office suite
Microsoft 365 Copilot
The AI inside Word, Excel, Outlook and Teams. It can reach whatever the signed in person can reach.
Verified 2 July 2026 →
Google Workspace with Gemini
Gemini inside Gmail, Docs, Drive and Meet. Google says your content is not used to train models outside your domain. What it can reach is the longer answer.
Verified 3 July 2026 →AI chatbots
ChatGPT
The one your staff are probably already using. What happens to what they type depends entirely on whose account it is.
Verified 2 July 2026 →
Google Gemini
Google's chatbot, one tap away on Android phones, in Chrome and in search. On a personal account, the training switch is on until someone finds it.
Verified 2 July 2026 →
Claude
Anthropic's chatbot. The same personal versus organisational story as the others, with the training question put to you as a setting.
Verified 2 July 2026 →Meeting notetakers
Otter.ai
The notetaker that joins your meetings, and can invite itself to the next one. What it hears may help train its AI.
Verified 2 July 2026 →
Microsoft Teams
The meeting recorder your charity already owns. Everything it captures becomes a file in OneDrive and SharePoint, where the rest of the AI can reach it.
Verified 3 July 2026 →
Zoom AI Companion
The AI inside the video calls much of the sector runs on. Zoom's no training promise is unusually specific, and there is a reason it is worded that way.
You are here
Fathom
The free notetaker that arrives one personal install at a time. The recordings live in the US, and the training answer has two halves.
Verified 3 July 2026 →Charity CRMs
Beacon
A UK charity CRM with AI on every plan since 2024. Beacon publishes where your data sits and how to switch the AI off, which leaves one question to ask them directly.
Verified 2 July 2026 →
Donorfy
A UK charity CRM where the AI mostly arrives through the side door: optional integrations. What you connect decides where your supporter data goes.
Verified 2 July 2026 →
CiviCRM
The open source CRM ships no AI of its own. What that is worth to you depends on which CiviCRM you are running: one your organisation hosts, or a hosted product somebody sells you. Everything below splits along that line.
Verified 20 July 2026 →
Salesforce
In much of the sector through ten free licences. The AI paperwork is detailed and mostly reassuring, with one default that runs the other way.
Verified 3 July 2026 →Accounts
Xero
Your accounts package now has an assistant called JAX. Xero says your data is not used to train its AI, and names the AI companies your questions travel to.
Verified 20 July 2026 →
QuickBooks
Intuit's accounts package now has an assistant throughout. Its privacy statement says your information is used to train its AI models, and it says so in as many words.
Verified 3 July 2026 →Marketing and design
Mailchimp
Your supporter mailing list lives here, and Mailchimp belongs to Intuit now, which means the same small print as QuickBooks covers it.
Verified 3 July 2026 →
Canva
The design tool half the sector lives in, free through Canva for Nonprofits. Whether your uploads help train AI depends on whose account they sit in.
Verified 3 July 2026 →