Otter AI accused of secretly recording private work conversations
Otter AI accused of secretly recording private work conversations
Occurred: 2025
Page published: September 2026
Popular AI meeting-transcription tool Otter.ai stands accused of secretly recording and transcribing private conversations, including those of non-users who never consented, and using the data to train its AI models, raising serious privacy, consent, and corporate accountability concerns.
A class-action lawsuit filed in northern California accuses Otter.ai of failing to ask attendees for permission to record their meetings and fails to alert participants that recordings are shared with Otter to improve its AI systems.
The lead plaintiff, Justin Brewer of San Jacinto, California, alleges his privacy was "severely invaded" upon realising Otter was secretly recording a confidential conversation. Notably, the plaintiff wasn't even an Otter user; his conversation was captured simply because another participant had the tool running. Several other plaintiffs later joined a consolidated case.
The system impacted both registered users and non-subscriber third parties who never agreed to Otter's terms of service by intercepting and exploiting sensitive medical, financial, and professional discussions.
The issue stems from Otter’s architectural design, which relies on automated bots joining scheduled calendar events as "silent participants," alongside data retention practices that store audio and transcripts for model training.
Transparency and accountability limitations compounded the issue: Otter’s business model attempted to shift legal compliance obligations onto meeting hosts (requiring them to ensure permissions) while omitting clear real-time consent pop-ups or mandatory opt-ins for non-subscriber attendees on the call.
For individuals and professionals, the controversy highlights severe digital eavesdropping and data harvesting risks inherent in ubiquitous workplace AI productivity tools.
For policymakers, the case highlights a regulatory gap: privacy and wiretap laws written before ubiquitous AI transcription didn't anticipate one participant's software silently drawing in everyone else's speech for model training. It also strengthens the case for stronger, clearer consent requirements specific to AI recording tools, and for treating "training data" use as a distinct, separately-consented purpose from basic transcription.
System: Otter.AI
Developer: Otter.ai Inc.
Deployer: Justin Brewer
Purpose: Transcribe work conversations
Technology: Audio-to-text; Machine learning; NLP/text analysis; Speech recognition
Ethical issue: Accountability; Consent; Privacy/surveillance; Transparency
External harm: Privacy loss
Impacted stakeholder: End user
Impacted sector:
Jurisdiction: California, USA
Consequence: Litigation
Response:
Justin Brewer v Otter.ai
AIAAIC Repository ID: AIAAIC2276