AI Note-Taking Apps Are Violating User Privacy and Facing Lawsuits

A few years ago, automated transcription and meeting note tools served as niche productivity hacks primarily utilized by media professionals and busy executives. Today, they have become ubiquitous in daily workflows, yet these applications come with legal fine print that the average consumer rarely examines before hitting record.

Primarily, these apps may retain your meeting recordings and transcripts indefinitely after a session concludes. Furthermore, they reserve the right to repurpose this data for various internal operations, including marketing campaigns and machine learning model training. Compounding the issue, state regulations frequently mandate that all conversation parties must consent prior to recording, whereas AI transcription developers contend that the legal obligation to secure such consent falls entirely on the individual operating the software—meaning you.

Stock photo for illustration only, not from the actual event

Consequently, multiple artificial intelligence note-taking platforms now face class-action lawsuits brought by users accusing them of mishandling sensitive data and shifting legal compliance burdens onto customers without adequate disclosure. For instance, Granola's design ensures that meeting participants cannot disable the software except for the individual host operating the device. Additionally, users are automatically opted into model training by default, requiring manual opt-outs, and opting out retroactively does not purge prior conversation records already captured for training.

This scrutiny extends beyond a single firm, as Silicon Valley witnesses multiple active legal actions against automated note-takers. In 2025, four separate class-action complaints against Otter.ai were consolidated before Judge Eumi K. Lee in the U.S. District Court for the Northern District of California. Similar to Granola, the litigation asserts that Otter records interactions without participant authorization to train proprietary models. Justin Brewer, a plaintiff in the Otter proceedings, testified that his voice was captured via OtterPilot during a sales discussion and subsequently repurposed for AI training without his consent or awareness. Although defense attorneys have filed motions for dismissal, the presiding judge has yet to issue a final ruling.

To safeguard your privacy and maintain compliance while utilizing these tools, consider implementing the following proactive steps:

• Obtain explicit verbal consent from every meeting participant prior to initiating any recording session.
• Treat meetings involving residents of California, Illinois, or other all-party consent jurisdictions with strict verbal permission protocols.
• Inquire directly with the meeting host if an AI transcription tool or secret recorder is active when no prior disclosure was provided.
• Navigate application settings beforehand to actively opt out of any data sharing or model training agreements.
• Invoke regional privacy rights, such as California's right to erasure, by submitting written notices to customer service requesting the deletion of legacy data.

Source: Lifehacker