That voice recorder app in your meetings isn’t just taking notes. It’s storing your conversations, analyzing them, and in some cases, feeding them back into AI models without you realizing it. The class action lawsuits filed against Granola, Otter.ai, and others should make anyone who uses these tools pause.
Your meeting notes app is probably not as private as you think
Here’s the thing most people miss: when you join a Zoom call and an AI note-taker starts transcribing, that audio doesn’t just vanish after the summary is generated. It gets stored. Sometimes indefinitely.
The problem goes deeper than storage, though.
What these apps actually do with your data
Most AI note-taking platforms process your meeting audio through their servers. That’s where the transcription happens, and that’s where the data lives. But “processing” is a broad term. It can mean:
- Storing raw audio recordings for months or years
- Keeping transcripts in searchable databases
- Using your conversations to train and improve their AI models
- Sharing aggregated insights with third parties (anonymized, supposedly)
The exact details vary by platform. Some apps let you opt out of model training. Others bury that option three clicks deep in a settings menu you’ll never visit.
The lawsuits happening right now
This isn’t speculation. Multiple class action lawsuits have been filed in 2026 against AI note-taking platforms, with users accusing them of mishandling data and failing to get proper consent. The core complaint? These apps didn’t clearly tell users that their private business conversations could be used for AI training.
There’s a secondary issue too. In many US states, recording a conversation requires consent from everyone involved. AI note-taking apps argue that responsibility falls on you, the person who hit record. But most users had no idea that was their legal obligation.
Sound familiar? If you’ve ever used an AI note-taker without asking everyone on the call first, you might be more exposed than you think.
Which AI note-taking apps have the biggest privacy problems
Not all note-taking apps handle data the same way. Here’s a breakdown of the ones currently in the spotlight and what you need to know.
Granola
Granola became a favorite among professionals for its clean interface and smart summaries. But it’s also one of the apps named in privacy lawsuits. The complaint centers on how Granola handles meeting data, specifically whether users were adequately informed about data retention and model training policies.
If you use Granola, check your account settings. Look for options related to data retention, model training opt-out, and audio storage. Delete old transcripts you no longer need.
Otter.ai
Otter.ai has been around the longest and has one of the largest user bases. That also makes it a bigger target. The platform stores meeting transcripts and, depending on your plan, may retain audio recordings. Users in the lawsuit argue that the consent mechanism wasn’t clear enough, especially for people who joined meetings where someone else was running Otter.
Otter does offer some privacy controls. You can delete individual conversations, adjust data retention settings, and manage team-wide permissions. But again, these options aren’t front and center.
Other apps to watch
The lawsuits have primarily targeted Granola and Otter.ai, but the same concerns apply broadly. Any AI note-taking app that processes audio on remote servers carries similar risks. This includes newer entrants like Fireflies.ai, tl;dv, and meeting-specific features built into Zoom and Microsoft Teams.
The pattern is consistent: convenient AI features with privacy policies that most users never read. If an app is recording your calls, assume the data is being stored somewhere until you verify otherwise.
How to protect your meeting data right now
You don’t have to stop using AI note-taking apps. But you should use them with your eyes open. Here are practical steps you can take today.
Check your privacy settings today
This takes five minutes and could save you a headache later. In whichever app you use, navigate to settings and look for:
- Data retention period — how long are recordings and transcripts kept?
- Model training opt-out — is your data being used to train their AI?
- Audio storage — is the actual audio file saved, or just the text?
- Sharing defaults — who can access your meeting notes by default?
Change anything that doesn’t feel right. If an app doesn’t let you control these settings, that’s a red flag.
Delete old transcripts you don’t need
Most people accumulate hundreds of meeting transcripts over time. Each one is a potential data exposure point. Go through your history and delete anything that contained sensitive information: financial discussions, client data, personal details, strategy conversations.
This isn’t paranoid. It’s basic data hygiene.
Use local alternatives when privacy matters
For meetings where confidentiality is critical (legal discussions, HR matters, financial planning), consider running transcription locally. Tools like MacWhisper or the open-source Whisper model can transcribe audio on your device without sending anything to a server. The tradeoff is accuracy — cloud-based models are still better — but for sensitive conversations, the privacy tradeoff is worth it.
Another option: use built-in OS features. Both macOS and Windows now have speech-to-text capabilities built in. They aren’t as polished as dedicated AI note-takers, but they keep your data on your machine.
And here’s the simplest rule: always tell people on the call that an AI tool is recording. It’s not just good manners. In many jurisdictions, it’s legally required.
The bottom line on AI note-taking apps and privacy
AI note-taking apps are genuinely useful. They save time, create searchable records, and help you focus on conversations instead of frantic typing. But that convenience comes with real privacy tradeoffs that most users never think about.
The lawsuits against Granola and Otter.ai are a wake-up call. Check your settings, delete old data, and use local alternatives when it matters. Your meeting notes shouldn’t become someone else’s training data.
If you want to explore more AI tools for meetings, check out our breakdown of the best AI meeting notes tools or see why tools like Screenpipe raise similar privacy questions. And if you’re concerned about AI and data exposure in general, our piece on the Claude privacy leak is worth a read.
