In August 2025, Otter.ai faced a federal class-action lawsuit alleging it 'deceptively and surreptitiously' records private conversations without proper consent. The lawsuit claims OtterPilot joins meetings automatically without obtaining consent from all participants and uses recordings to train AI models without permission.

Users report Otter automatically joins meetings without permission and sends transcripts to participants - a blatant privacy overstep. Once connected to a calendar, it's difficult to control when it will show up. The bot joins without obtaining affirmative consent from any participant.

The class action lawsuit alleges Otter used recorded conversations to train its AI models without explicit user consent. The complaint notes that de-identification is imperfect, particularly with voice data and conversational context, raising serious data privacy concerns.

The number one complaint in Otter.ai reviews is accuracy issues. Transcripts often top out around 85% accuracy, requiring frequent edits. Users always need to edit everything and almost every sentence. If participants have strong accents, use industry jargon, or talk over each other, transcripts get messy.

A common headache is Otter's trouble telling speakers apart. Transcripts often end up with a sea of 'Speaker 1, Speaker 2' making it impossible to figure out who said what. This messes up summaries and action items too. Speaker names from meetings often aren't detected.

Otter rarely captures the real 'next steps' or key takeaways from meetings, meaning users still have to manually review everything. In meetings with layered topics or fast-paced discussion, Otter produces generic summaries that lack depth. It transcribes what was said but not the why or how.

Otter handles standard English well but much less so African, Arabic, Somali, or non-native French accents. International teams and diverse workplaces find transcription quality drops significantly when participants have various accents, making the tool unreliable for global organizations.

Notes emailed to members after meetings are only accessible with an Otter account, which is not always suitable when meeting with those outside the organization. External clients or partners need to create accounts just to view shared transcripts, creating friction.

Otter.ai has been charging accounts $30/month without sending receipts or billing notifications. Users canceled immediately yet charges continued. Some selected monthly but were billed for the whole year without receiving responses to refund requests. 'Shrinkflation' gives fewer minutes for the same price.

Users report confusing cancellation steps and frustration with free trials that didn't allow full meeting recordings. Others report billing problems and unexpected charges. Some selected monthly plans but were billed annually with no response to refund requests.

The free plan comes with only 300 minutes per month, 30-minute max per conversation, and just three file uploads for life. This is where most Reddit frustration originates. No OtterPilot auto-join on free tier means manual recording, and no team sharing features.

Many users have pointed out 'shrinkflation' where they get fewer minutes for the same price they used to pay. Having to track usage and stay within transcription minute limits adds unnecessary stress to workflow. Heavy usage months can lead to surprise costs or getting cut off.

There is no customer support if a problem comes up. Reviewers describe customer service as non-existent with no replies or help. Response times are slow, sometimes taking days to get a reply. This is especially problematic for billing-related inquiries which add to user frustration.

Some users say Otter fails to capture transcripts reliably. The tool sometimes misses portions of meetings or fails to join altogether. Time codes are not always accurate. Users recommend alternatives like Fathom or Avoma which they find more consistent and better value.