Glossary
False accept (accidental activation)
A false accept, also called a false wake or accidental activation, is when a voice assistant or dictation feature starts listening or recording although the wake word was not said. Smart speakers have been measured to wake by accident, and that can capture speech the user did not mean to share.
Updated October 17, 20264 cited facts, each traced to its source
Also called false wake, accidental activation, false trigger.
False accept is what happens when a voice system decides that its wake word was spoken, and starts listening or recording, when it was not. It is also called a false wake or accidental activation. The opposite mistake, a missed wake word, is sometimes called a false reject. Every wake-word detector trades these two errors against each other, so a lower chance of missing a command usually means a higher chance of a false accept.
A wake-word detector listens all the time
A wake-word detector runs a small model on sound continuously, looking for a short phrase. Microsoft says that its "Hey Copilot" wake word is detected on the PC by a small on-device service that keeps a 10-second audio buffer in memory to listen for the phrase, and that the buffer is never recorded or stored 1. The design shows the basic problem. The detector must judge every sound against one phrase, so sounds that resemble the phrase can pass, even when no audio is kept.
Measured rates: smart speakers on TV audio
In a 2020 study of smart speakers exposed to TV shows, a Google Home Mini playing The West Wing averaged 0.95 activations per hour without the wake word, which the authors singled out as a notable case 2. The authors found no evidence that the devices constantly record conversations, but they did find that the devices wake up frequently, often for short intervals 3. Those figures describe the devices and shows tested in 2020, and newer models may behave differently.
A concrete example: a recording that starts early
Google says that Gemini apps wait in standby until activated, and that audio recordings may begin a few seconds before activation 4. A recording that begins before activation can include words that were never meant for the device.
A dictation example is hypothetical. A dictation feature left listening during a television show picks up a phrase that sounds like its trigger, opens a session, and types the show's dialogue into an open document. You would only see it afterward, which is why the trigger design matters.
Why it matters for dictation
Dictation that listens all the time, such as a hands-free wake phrase, brings false accepts into the text itself and may also capture audio. A push-to-talk key or a visible microphone indicator makes activation explicit, so a false accept is less likely to start a session without you noticing. Choose the trigger that fits the room: where others talk or a television plays, a manual trigger means no session starts unless you start one.
Check the settings too. Look for the wake phrase option, whether audio is stored after a session, and whether the microphone indicator shows when the device is listening. Our guide explains how often voice assistants activate by accident, and what to check on your own devices.
Related terms
- Wake word: the phrase that a detector listens for before it starts a session.
- Voice activity detection (VAD): the component that decides whether audio contains speech.
- Push-to-talk: a trigger that starts listening only while you hold a key.
Guides and questions that cover it
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.In a 2020 study of smart speakers exposed to TV audio, the authors found no evidence that the devices constantly record conversations. They did find that the devices wake up frequently, usually for short intervals.
Daniel J. Dubois et al., Proceedings on Privacy Enhancing Technologies (PETS 2020), 2020 · Original
- 4.
Related terms
- Speech to Text Privacy: Where Your Voice GoesThe topic this term belongs to
- All-party consentAll-party consent is a recording rule under which every participant in a conversation must agree before it can be recorded or intercepted. California, Florida and Maryland use it in their statutes, and it is stricter than one-party consent.
- Attorney-client privilegeAttorney-client privilege protects confidential communications between a lawyer and their client about seeking legal advice or services. A third party who is present, or a tool that handles the conversation, can put that protection at risk.
- Biometric dataBiometric data is personal data produced by technical processing of a person's physical, physiological or behavioral characteristics that allows or confirms who that person is. A voice recording becomes biometric data when it is processed to identify the speaker.
- Business associate agreement (BAA)A business associate agreement (BAA) is the written contract that HIPAA requires before a covered entity lets a vendor create, receive, maintain or transmit protected health information on its behalf. Dictation tools used in clinics often need one, and a BAA alone does not make a product compliant.
- Federated learningFederated learning is a way to train one machine learning model on data spread across many devices. Each device computes an update from its own data, and only those updates are sent to a server, so the raw data does not have to be collected in one place.
- On-device speech recognitionOn-device speech recognition converts speech to text with a model that runs on your own computer or phone, so the audio does not need to be sent to a server. It is also called local, offline or on-device dictation.
- One-party consentOne-party consent is a recording rule under which a conversation can be recorded lawfully when one of its participants agrees, which is usually the person doing the recording. The US federal wiretap law uses this rule, and individual states may set stricter ones.
- Protected health information (PHI)Protected health information (PHI) is individually identifiable health information that HIPAA protects when it is held or sent in electronic or other form by a covered entity or its business associate. Patient names with diagnoses, in dictated notes or voice recordings, are PHI. It is also called ePHI when held electronically.
- Speaker verificationSpeaker verification is a voice biometric check in which a system accepts or rejects a claimed identity by comparing a voice with stored models of the person claimed. It is also called voice authentication, voice ID or voice biometrics.
- Special category data (GDPR)Special category data is the group of personal data that Article 9 of the GDPR bans organizations from processing unless a listed exception applies. It includes health data and biometric data used to identify a person, and a voice recording can fall into it when it is processed that way.
- Voice cloningVoice cloning is the use of a model trained on recordings of one person to generate new speech in that person's voice. Some products need only minutes of clear audio, and cloned voices have been used in fraud, so recordings of your voice carry a real risk.
- VoiceprintA voiceprint is a stored model of a person's voice, built from measurable features of their speech and used to recognize or verify that person. Illinois law lists voiceprints among biometric identifiers, and other vendors call the same idea a voice template or speaker model.