Question · Voice privacy and security
Why do companies ban AI notetakers and dictation apps?
Companies ban a tool when the vendor, not the company, controls what happens to what people say, or when recording breaks consent rules. Personal accounts add risk because they sit outside company logging and retention controls. The public bans we found target chatbots and a meeting notetaker, not dictation apps by themselves.
Updated October 17, 20263 min read7 cited facts, each traced to its source
Companies ban a tool when they cannot control what the vendor does with the audio or the text. Most of the bans we found concern AI chatbots and meeting notetakers. Dictation apps usually fall under the same test: where the speech goes, how long it stays, whether it trains a model, and who has the account. A notetaker adds a second problem, because it records people who never chose the vendor.
A ban usually follows loss of control over the audio
The first concern is where the data goes. Samsung's internal memo said the company was worried that data sent to AI platforms is stored on external servers, where it is hard to retrieve and delete 1. Samsung confirmed that it was temporarily restricting generative AI use on company computers after misuse of the technology 2.
Dictation raises the same question through a different input. Once speech leaves the device, the company loses direct control over it. A transcript is also a document, so it carries the same secrets as anything typed.
Personal accounts are the usual gap
Many of the risks come from the account rather than the tool. Cyberhaven, a data security vendor, says personal AI accounts bypass SSO enforcement, centralized logging, enterprise retention policies and controls on model training 3. A policy that approves a tool but allows any login leaves those controls behind. The usual fix is a rule that work content goes only through a work account, which is also how a company gets its logs and retention settings.
Notetakers draw the strictest response because they record other people
A notetaker that joins a call records people who did not choose its vendor, and those people may not know they are being recorded. A federal class action filed in August 2025 alleges that Otter.ai's Otter Notebook service, by default, does not ask meeting attendees for permission to record 4. These are allegations, and this page does not report a ruling on them. The vendor's own terms matter too. Otter's privacy policy, effective June 16, 2026, lists training its AI on de-identified audio recordings and on transcriptions among its purposes 5.
Consent law is the other reason. In California, Penal Code 632 punishes a person who intentionally records a confidential communication without the consent of all parties 6. Federal law is looser. It allows interception when the interceptor is a party to the communication, or when one of the parties has given prior consent 7. Whether a notetaker that joins a call counts as a party, and whose consent counts, depends on the state. Companies ban the tool rather than ask each employee to work that out, and our answer on whether an AI notetaker needs everyone's consent explains the state rules.
It depends on the contract, the settings and the call
A ban is one of several possible answers. The same tool can be treated differently depending on what it does:
- A tool with a signed contract and no training on company speech is often approved, because the contract answers the audit question.
- An on-device tool that never sends audio needs a settings and network check rather than a ban. The glossary entry on on-device speech recognition explains what that means.
- A consumer notetaker on client or patient calls is often banned even with consent, because the terms of a consumer account do not fit confidential work. For legal work, see our answer on lawyers and AI notetakers.
What to do
- Ask what the tool keeps. Find out whether it stores audio, text or both, who reviews them, and whether they can train a model.
- Require a work account for any cloud tool. Block personal accounts for work content, because that is where the controls disappear.
- Set a consent rule for notetakers. Decide before the meeting whether the notetaker joins at all, and check the consent law in each state where participants are located. The glossary entry on all-party consent defines the rule that applies in the strictest states.
- Write the decision down. Put the approved and banned tools in your dictation policy, with an owner and a review date. Our answer on what a dictation policy should include gives an outline.
For more, read our guide to dictation at work, enterprise and government policies and our overview of voice privacy. Our answer on whether your employer can see what you dictate covers the device and account side of the same question.
Evidence
Every number on this page, with where it comes from. Open a fact to see the exact words in its source.
- 1.
- 2.
- 3.
- 4.
- 5.
- 6.
- 7.
Go deeper
Related questions
- Can government employees use AI dictation?
- Can speech to text run on an air-gapped computer?
- Can you use speech to text for confidential documents?
- Can your employer see what you dictate?
- What should a company dictation policy include?
- Are conversations with an AI chatbot privileged?
- Are open-source offline dictation apps like Handy and Murmur safe to use?
- Can a third-party keyboard see what you type?
Referenced by
- Speech to Text Privacy: Where Your Voice GoesTopic
- Dictation at Work: Enterprise and Government PoliciesGuide
- Can government employees use AI dictation?Question
- Can your employer see what you dictate?Question
- Is it legal to use an AI notetaker without telling everyone?Question
- What should a company dictation policy include?Question