Otter AI Review (2026): The Bot That Joins Every Call, and the Lawsuit It Brought

This Otter AI review starts with an odd fact. The same week you invite Otter into a call, a federal judge is deciding whether that invitation counts as a wiretap. Otter has 35 million users and 100 million dollars in yearly revenue. It also has a class action that survived a motion to dismiss in August 2026.

Otter.ai is an AI meeting assistant from Mountain View, California. Sam Liang and Yun Fu founded it in 2016 under the name AISense, and it became one of the first transcription apps to plug straight into Zoom.

Today the product is a notetaker bot plus a knowledge base. The bot joins your Zoom, Google Meet or Microsoft Teams call, writes a live transcript and labels each speaker. Afterwards it hands you a summary, action items and a chat window over every meeting you ever recorded.

The company now calls itself a Conversational Knowledge Engine. Behind the label sits a simple bet: individual notes are not the product anymore, the searchable archive is.

Otter works on a simple loop. You connect your calendar, the bot shows up for scheduled calls, and the notes land in your workspace a minute after the meeting ends. Below are the four parts that shape daily use.

The Otter Notetaker appears as a participant named after you. Everyone in the call can see it, and it records the audio for the whole session.

That visibility cuts both ways. On one hand, nobody can later claim they were recorded in secret. On the other hand, a bot in the participant list changes how clients and candidates talk, and some companies simply block it at the door.

This is where Otter still leads. The live transcript scrolls while people talk, and speaker identification works on every plan, free included. Pro adds team vocabulary, so product names and surnames stop coming out mangled.

In addition, the audio stays attached to the text. Click a sentence and Otter plays that moment back, which settles most arguments about who promised what.

AI Chat answers questions within one meeting or across all of them. You can ask what a client said about budget last month, or list every open task for one colleague, without scrolling a single transcript.

The April 2026 rebuild made chat private by default and gave it a layout close to Claude and ChatGPT. Meanwhile the voice-activated Meeting Agent answers the same kind of question out loud, in the middle of a live call.

Otter now works as both an MCP server and an MCP client. As a server it lets Claude or ChatGPT read your meeting history. As a client it pulls Gmail, Google Drive, Notion and Salesforce data into Otter’s own chat.

Otter for Desktop closes the other gap. It captures audio on Mac and Windows from any app, including Slack Huddles, so a meeting no longer needs a bot to be transcribed.

After a stretch of daily use, the strengths are easy to name. Transcripts are accurate for clear English audio, speaker labels hold up in calls with five or six people, and the summaries arrive fast enough to paste into a follow-up email.

The weak spots this Otter AI review kept hitting are just as consistent. Accents, cross-talk and poor microphones still produce garbled passages. Summaries also lean generic, so the key decision in a meeting can end up as one bullet among twelve.

The auto-join behaviour is the bigger issue. Connect a calendar, leave the default on, and the bot joins every meeting on it, including ones you forgot about. That habit is exactly what the lawsuit below targets, and it is why so many people search for a way to remove it.

Otter has four plans, and every price in this Otter AI review comes straight off the official pricing page. Annual billing roughly halves the Pro price.

So is Otter AI free? Yes, but the limits are tight. Basic gives you 300 transcription minutes a month, caps each meeting at 30 minutes and allows only three file imports for the lifetime of the account. Pro raises that to 1,200 minutes, 90 minutes per meeting and ten imports a month, plus unlimited storage and the Salesforce, HubSpot and Zapier integrations. Business removes the meeting cap entirely, allows four-hour sessions and lets the bot join three meetings at once. Enterprise adds SSO, the API and webhooks, while HIPAA compliance is a paid add-on rather than a default.

In August 2025 four proposed class actions accused Otter of recording meetings without consent. The plaintiffs said the Otter Notetaker captured Zoom, Meet and Teams calls, including people who never signed up, and used that audio to train its AI models.

The cases were merged as In re Otter.AI Privacy Litigation in the Northern District of California. In August 2026 Judge Eumi K. Lee refused to throw the case out, as Bloomberg Law reported. Claims under the federal Wiretap Act, the California Invasion of Privacy Act and Illinois’ biometric privacy law all go forward.

Some counts were dismissed with leave to amend, and nothing has been proven yet. However, the ruling means the case moves toward discovery and class certification. For you, the practical point is consent: if you run Otter, tell the room, because the other people on the call never agreed to Otter’s terms.

The fastest fix happens inside the meeting. Remove the Otter participant the way you would remove any guest, and the recording stops for that call.

To stop it for good, sign in to Otter and open Account Settings, then the Meetings tab. Switch off auto-join for all meetings, or set it to manual. Next, disconnect the calendar Otter is reading, so it cannot see future invites.

For a clean break on Zoom, open the Zoom App Marketplace, go to Manage and remove Otter from your added apps. On Teams, a workspace admin can block the notetaker for the whole organisation, which is the only fix that also covers guests who bring their own bot.

Otter is a legitimate, established company, and the facts behind this Otter AI review back that up. It has a named founder and CEO in Sam Liang, fewer than 200 employees, and it reported crossing 100 million dollars in annual recurring revenue in 2025. It holds SOC 2 Type II and added HIPAA compliance in July 2025.

The caveats concern data, not legitimacy. Your recordings live on Otter’s servers, the auto-join default is aggressive, and the company’s use of meeting audio for training is now in front of a federal court.

In short, Otter is safe to use as a product, provided you control the settings. Turn off auto-join, announce the bot, and check your own company’s recording policy before a sensitive call.

Granola is the opposite philosophy. No bot joins, nothing appears in the participant list, and your own typed notes drive the summary. Our Granola AI review covers it in depth.

Choose Granola when a visible recorder would change the conversation. Stay with Otter when you need replayable audio and reliable speaker labels.

Fathom offers the most generous free tier in the category, with unlimited recording and transcription. It also clips highlights and pushes notes into a CRM with little setup.

Therefore it beats Otter’s 300-minute free plan for anyone on calls all day. Otter keeps the edge on live transcripts and cross-meeting chat.

Fireflies.ai targets teams. It records, scores calls and builds a searchable archive across the organisation, backed by a long list of integrations.

As a result it suits sales and support managers who review other people’s calls. A single user will find Otter quicker to set up and easier to read, the verdict this Otter AI review gives most solo buyers.

Pros

• Live transcript with speaker labels on every plan, free included
• Click any sentence to replay the original audio
• AI Chat and the Meeting Agent search across your whole history
• MCP server and client connect meetings to Claude, ChatGPT, Gmail and Notion
• Otter for Desktop captures calls without a bot on Mac and Windows
• Pro costs 8.33 dollars a month when billed yearly

Cons

• Auto-join sends the bot into every calendar meeting unless you switch it off
• An active wiretap class action over consent and AI training
• The free plan caps meetings at 30 minutes and allows 3 lifetime imports
• Accents, cross-talk and weak microphones still produce errors
• Summaries can bury the key decision among generic bullets
• HIPAA compliance costs extra, even on Enterprise

Otter remains the most complete meeting recorder for most people. The transcripts are strong, the replay is useful, and the new chat and MCP features turn a pile of notes into something you can actually query.

Still, the product asks for trust it has not fully earned. The auto-join default and the training question are the reasons a judge is involved, and neither is a problem you can outsource to the vendor.

So here is the short answer from this Otter AI review. Pay for Pro if you need recordings, speaker labels and a searchable archive, and set it up deliberately. If what you want is a quiet page of notes, Granola will serve you better. If you want free and unlimited, start with Fathom.

We ranked the whole category, bots and notepads side by side, in our guide to the best AI note taker tools. And if your notes should trigger tasks on their own, see the best AI workflow automation tools.