This post is also available in:
עברית (Hebrew)
A widely used AI-powered transcription service is under legal scrutiny in the United States over concerns that it may be collecting and using recorded conversations without proper consent. The platform in question, known for its automatic meeting transcription tool “Otter Notetaker,” is accused of using recorded audio data to train its speech recognition algorithms.
According to The Register, the issue came to light in a recent lawsuit filed by a user who claims that their conversations were captured and repurposed by the service without adequate notice. At the center of the complaint is the allegation that Otter’s system records and processes everything said during a meeting, and then uses that data to refine its AI models, regardless of whether participants are account holders or not.
According to the lawsuit, Otter places the responsibility for consent on the user who initiates the recording, rather than seeking direct approval from all attendees. This raises questions about compliance with privacy regulations, particularly when sensitive or proprietary information is being shared in virtual meetings hosted on platforms like Zoom, Microsoft Teams, or Google Meet.
The transcription tool is designed to record meetings, and may do so automatically, without alerting or receiving explicit permission from participants. In many cases, the software may begin recording by default, especially if integrated directly into conferencing platforms. This has led to concerns that users may not even be aware when the software is active, or that their conversations and voices are being stored and analyzed.
The implications could be significant, particularly for businesses. If sensitive internal discussions or confidential data are being collected and reused by an external service provider, companies may face compliance issues, contract breaches, or unintended data exposure.
As legal proceedings move forward, the case highlights the growing challenges in balancing AI development with data privacy, especially in environments where recording tools operate passively in the background.
