Granola Sued for Secret Meeting Recordings, Raising Privacy Alarms Across AI Note-Taking Industry
Summary
- • AI note-taking app Granola faces a class action filed July 30 in California federal court alleging it secretly records calls without all participants' consent, violating CIPA — California's all-party consent law
- • Granola captures audio directly from the user's computer without appearing as a meeting bot, making it invisible to other call participants; the lawsuit alleges the company markets this hidden capability as a feature advantage
- • By default, Granola uses transcription data to train its AI models — a practice the suit argues was done without meaningful informed consent from all parties recorded
- • A parallel Otter.ai class action is before the same California district court, where a judge recently expressed skepticism about dismissal arguments — two cases are setting legal precedent for the entire AI note-taking industry
Details
Class action filed July 30
Plaintiff Tarra Chamberlain (Florida resident) filed in US District Court for the Northern District of California; alleges violations of CIPA (all-party consent) and individual privacy rights
OS-level audio capture
Granola captures audio directly from the user's computer without appearing as a meeting participant — unlike bot-based tools — making recording invisible to all other call participants
Hidden recording marketed as feature
Complaint alleges Granola 'purposefully' designed the app to record without disclosure and 'actively advertises the hidden nature of its technology as one of its primary advantages'
AI training use alleged
Transcription data is used by default to train Granola's AI models for commercial purposes; company says data is anonymized and never sent to third parties, but consent of all recorded parties was not obtained
Optional transparency features exist
Granola offers two opt-in transparency features: an automated chat alert when transcription begins, and a watermark on the user's video — but both must be manually enabled by users or admins
Parallel Otter.ai case
A separate class action against Otter.ai is before the same district court; at an August 4 hearing, Judge Eumi K. Lee expressed skepticism about Otter's dismissal arguments; written ruling pending
Forrester analyst on AI-specific risks
Forrester VP Enza Iannopollo: AI note-taking is 'more dangerous than any other type of traditional recording' because it raises additional questions about biometric data, model training, and the right to be forgotten after recording
Granola faces a class action over covert AI meeting recordings and consent-free AI training data collection, mirroring an ongoing Otter.ai case. Source: Computerworld AI.
What This Means
The Granola lawsuit is the second major CIPA-based challenge to an AI note-taking app in the same California district court, and together these two cases are beginning to define the legal boundaries of consent-based recording in the AI era. The core design choice at issue — whether capturing audio at the OS level without a visible bot constitutes secret recording under all-party consent laws — will determine the legality of an entire product category used by tens of millions of workers. Enterprises that have already deployed these tools may face secondary exposure, and the cases are likely to push AI note-taking vendors toward mandatory transparency features or significant product redesigns. Businesses should audit their current AI note-taking deployments and ensure all contractual and disclosure requirements are in place before the legal landscape clarifies.