Otter.ai

Is Otter.ai safe to use for client meetings? A 2026 class action put the whole AI-meeting-bot category on notice — and recording other people carries obligations whichever tool you use.

Plain-English risk rating: 5 of 5 (Free/Pro pending litigation outcome) / 3 of 5 (Business/Enterprise)

Otter.ai is currently the highest-risk meeting AI tool in this database, and the risk picture is dominated by an active and consolidated class action lawsuit that goes to the heart of how Otter operates: whether OtterPilot's auto-join behaviour and voiceprint generation constitute unauthorised wiretapping and biometric data collection under federal and California law.

As of profile verification date (May 24, 2026), Otter is the defendant in In re Otter.AI Privacy Litigation (No. 5:25-cv-06911 and consolidated cases), a federal class action consolidated before Judge Eumi K. Lee in the Northern District of California. On 13 August 2026 the court ruled, granting the motion in part and denying it in part. The federal wiretap (ECPA), California Invasion of Privacy Act and Illinois biometric (BIPA) claims survive and proceed; the Computer Fraud and Abuse Act, California CDAFA and Washington Privacy Act claims were dismissed. Judge Lee rejected the argument that Otter was merely a tool acting for the meeting host, holding it could be treated as a third-party eavesdropper because it retained the conversations and used them for its own commercial purposes, including training machine-learning models. That was the first federal test of whether decades-old wiretap statutes reach an AI bot sitting quietly in a video call, and Otter largely lost it. The exposure for Otter, and for businesses using Otter, is now confirmed rather than uncertain.

The Business and Enterprise tiers carry the same underlying legal exposure but offer stronger contractual controls. The 3-of-5 rating for those tiers reflects the higher floor of vendor-side controls; it is not a guarantee that the litigation outcome will not affect those tiers.

Recommended for

  • Sole proprietor: Not recommended during the litigation pendency. ZoomMate is the closest privacy-comparable alternative. Granola was previously named here; that recommendation was withdrawn on 3 August 2026 following the class action filed against Granola on 30 July 2026.
  • Small team (2-10 people): Not recommended during the litigation pendency unless Otter is contractually required by an existing client relationship. Migrate planning is reasonable.
  • Regulated industry: Not appropriate. The biometric voiceprint generation alone is incompatible with regulated-data handling without extensive legal review and consent processes.
  • The honest answer for most 1-10 employee businesses: Otter.ai built the AI notetaker market and remains highly accurate, but its legal posture as at August 2026 remains the most exposed of any vendor in this database — though Granola has now joined it in facing a wiretapping-style class action. The right move is to pause new Otter deployments because the motion to dismiss was decided on 13 August 2026 and the core wiretap, CIPA and biometric claims survived, so the exposure is confirmed rather than speculative. Be careful about migration targets: Fathom and Granola both train their own models on de-identified meeting content by default, and Granola now faces its own class action. ZoomMate is the safer default, and whichever tool you use, turn training off and disclose recording. If you cannot migrate, the absolute minimum is to disable OtterPilot's calendar-driven auto-join entirely.

Critical pre-deployment warning (active class action litigation)

This is the most important fact in this profile: Otter.ai is currently defending an active class action that alleges its core product behaviour constitutes unauthorised wiretapping and biometric voiceprint collection without all-party consent.

The consolidated litigation includes four cases (Brewer, Walker, Theus, Winston), filed August through September 2025, consolidated by Judge Lee on October 22, 2025, with a consolidated complaint filed December 5, 2025. The core allegations:

  • OtterPilot (now rebranded as Otter Meeting Agent) syncs with the user's calendar and auto-joins any scheduled call as a visible participant
  • It records audio, transcribes in real time, takes screenshots, and captures speaker voiceprints
  • Non-account-holders attending those meetings receive no disclosure before the recording begins
  • Otter shifts consent responsibility onto the meeting host rather than obtaining it from all participants
  • Recordings and voiceprints are alleged to be used for AI model training, including for non-account-holders

Under California's Invasion of Privacy Act and analogous statutes in 11 other states (the "two-party consent" jurisdictions), all-party consent is required for recording. The Walker case specifically alleges that Otter captures and stores biometric voiceprints to identify the same speakers across future meetings — the BIPA equivalent of fingerprinting.

The practical implication for SMBs using Otter today: if a meeting participant in a two-party-consent state successfully sues over an Otter-recorded meeting, the responsibility may fall on the business that hosted the meeting (under Otter's contractual host-responsibility model) rather than on Otter itself. Document your consent practices carefully; review whether your client meetings could expose your business to liability.

Data retention default

  • Free, Pro, Business: Recordings and transcripts retained per plan; check current retention policy at otter.ai/privacy. Plaintiffs allege Otter retains recordings indefinitely.
  • Voiceprints (biometric): Allegedly retained for speaker recognition across future meetings; the exact retention policy is one of the contested issues in litigation
  • De-identified training: Otter states recordings are de-identified before being used for model improvement; this de-identification claim is contested by plaintiffs and is one of the litigation's core factual questions
  • Business and Enterprise: Admin-configurable retention; verify per current contract
  • OtterPilot agent activity: Calendar integration means OtterPilot may auto-join meetings; user-side controls available but the disclosure to non-account-holder participants is the litigation flashpoint

Training opt-out

Free and Pro — De-identified training is the default, with Otter asserting that recordings are de-identified before being used for model improvement. The de-identification claim is contested in litigation. Voice embeddings have demonstrated re-identification rates of >92% even from anonymised datasets (Nature Communications, 2023), which is one of the empirical arguments plaintiffs will likely make.

Opt-out controls have changed multiple times during 2024-2026; verify current settings via Account → Settings → Privacy. The individual opt-out availability is itself one of the regulatory questions.

Business and Enterprise — stronger contractual training exclusions available; verify per-contract.

Zero Data Retention availability

  • Not offered in the OpenAI/Anthropic API ZDR sense
  • Enterprise customers can negotiate custom retention terms

Plan tiers and pricing (as of early 2026)

TierPrice (USD)De-identified training defaultSuitable for
Free$0YesPersonal use; limited minutes
Pro$16.99/monthYesIndividual professionals (litigation exposure applies)
Business$30/user/monthVerify per contractSmall teams (litigation exposure applies but stronger contractual controls)
EnterpriseCustomNo — contractually excludedLarger orgs

Jurisdiction

  • Primary processor: Otter.ai Inc., Mountain View, California, USA
  • Cloud infrastructure: AWS
  • SOC 2 Type II certified
  • Subject to California privacy law (CCPA/CPRA), California Invasion of Privacy Act (the central litigation statute), federal Electronic Communications Privacy Act, federal Computer Fraud and Abuse Act, and BIPA (Illinois) for voice biometric data classification
  • Litigation venue: Northern District of California (consolidated cases before Judge Eumi K. Lee)

Breach history (public incidents and active litigation)

August 2025 onward — In re Otter.AI Privacy Litigation (ACTIVE CLASS ACTION)

The central event dominating Otter's risk picture as of mid-2026. Four federal class actions filed in California between August and September 2025, all alleging that OtterPilot's auto-join behaviour and voiceprint collection violate federal wiretap law, California's Invasion of Privacy Act, the Computer Fraud and Abuse Act, BIPA (in some cases), and common-law intrusion and conversion claims.

Key case timeline:

  • August 2025: Brewer v. Otter.ai Inc. (No. 5:25-cv-06911) filed in the Northern District of California. Plaintiff Justin Brewer was not an Otter account holder; he participated in a February 2025 Zoom call where Otter's notetaker auto-joined and allegedly created a voiceprint of him.
  • August 26, 2025: Walker case filed (voiceprint focus, biometric data allegations)
  • September 3, 2025: Theus case filed
  • September 10, 2025: Winston case filed
  • October 22, 2025: Judge Eumi K. Lee consolidates all four cases as In re Otter.AI Privacy Litigation
  • December 5, 2025: Consolidated complaint filed by interim co-lead counsel (Levin Law, Clarkson Law Firm, Werman Salas)
  • April 2026: MLex reports Otter's reply brief denying interception occurred
  • Motion to dismiss: decided on 13 August 2026. Judge Eumi K. Lee granted the motion in part and denied it in part, and the core claims SURVIVED: the federal wiretap (ECPA), California Invasion of Privacy Act, Illinois biometric (BIPA), unfair competition and unjust enrichment claims all proceed. Dismissed were the Computer Fraud and Abuse Act, California CDAFA, Washington Privacy Act, and the common-law privacy claims for every plaintiff except Theus. The reasoning matters more than the tally: the court rejected the argument that Otter was merely a tool acting for the meeting host and held it could be treated as a THIRD-PARTY EAVESDROPPER, because it retained the conversations and used them for its own commercial purposes, including training machine-learning models. Using the recording to improve your own product is what the court treated as the line between a participant and an eavesdropper. Corrected 2 September 2026: until today this profile said the motion was fully briefed and pending and that no ruling had issued. That was accurate when last checked on 3 August; the ruling landed ten days later and we did not catch it, because our monitoring watches vendors privacy pages and not court dockets. Verified against two independent reports of the order, both read 2 September 2026.

Otter CEO Sam Liang publicly defended the product to TechCrunch (October 7, 2025): "If they accuse us, then they could accuse everyone else, all the tools you heard about doing meeting notes. My view is that we are on the right side of history. We're building this new AI revolution. If you want AI to help, you need to put AI in the meetings."

Sources: court filings via PACER; Mason LLP analysis (April 2026); UC Today coverage (April 2026); Babst Calland analysis (January 2026); Fisher Phillips bulletin (August 2025); Best Law Firms analysis (November 2025); TLDV.io comparison guide (May 2026); HR Executive coverage (April 2026); Jackson Lewis P.C. coverage (August 2025); AI CERTs News analysis (February 2026)

December 2025 onward — Cruz v. Fireflies.AI Corp. (related BIPA litigation)

Katelin Cruz, an Illinois resident, filed a proposed class action in the Central District of Illinois, Springfield Division, against Fireflies.AI Corp. (No. 3:25-cv-03399-SEM-DJQ) on December 18, 2025. Cruz alleged that when Fireflies' bot joined a nonprofit meeting she attended, it recorded her voice and generated a voiceprint via Fireflies' "Speaker Recognition" feature without notice, written consent, or a retention policy as required by Illinois's Biometric Information Privacy Act. THE CASE IS OVER. Cruz voluntarily dismissed it on 10 March 2026 and Judge Sue E. Myerscough entered the order on 11 March 2026 under Rule 41(a)(1)(A)(i). Fireflies never filed a responsive pleading – it obtained two extensions and then the case was withdrawn – so nothing was decided either way and the allegations were never tested. Corrected 3 September 2026: this profile presented the Cruz case as live and placed it in the wrong court. It was filed in the CENTRAL District of Illinois, not the Northern, and it ended six months ago. Verified against the Justia docket for 3:25-cv-03399 and against the complaint itself, which was read in full: Katelin Cruz, filed 18 December 2025, pleading BIPA sections 15(a) and 15(b), alleging Fireflies recorded her voice and analysed her vocal characteristics to distinguish her as a unique speaker, and that it stores voiceprints of participants who never created accounts or agreed to its terms. Not Otter-specific, but directly relevant to the AI notetaker category and likely to inform the outcome of the Otter litigation.

Sources: court filings; Mason LLP analysis (April 2026); TLDV.io coverage

Note on institutional bans: Multiple universities have banned or restricted AI notetakers (Read AI, Fireflies, sometimes Otter) from Zoom and Teams integration during 2024-2026 (including the University of Washington, Chapman University, and the University of California, Riverside per UC Today). The pattern reflects institutional risk-management responses to the consent and biometric concerns.

No major direct breach of Otter's core infrastructure publicly disclosed as of May 2026. The dominant risk is litigation rather than infrastructure compromise.

What this means in plain English for SMB owners

Three honest takeaways:

  1. The active class action litigation is the dominant fact about Otter.ai as of mid-2026. Judge Lee’s ruling on the motion to dismiss came on 13 August 2026 and let the core claims through: federal wiretap, CIPA and BIPA all survive. The court held Otter could be treated as a third-party eavesdropper rather than the meeting hosts tool, because it kept the recordings and used them to train its own models. That reasoning reaches past Otter to any meeting tool that trains on what it records. The conservative posture is unchanged and is now better supported: pause new Otter deployments and plan migration of existing ones.
  1. The host-responsibility model shifts legal exposure onto your business. Otter's terms tell customers to "make sure you have the necessary permissions" — effectively outsourcing the consent compliance obligation. If you operate in a two-party-consent state (California, Washington, Florida, Illinois, Maryland, Massachusetts, Montana, Nevada, New Hampshire, Pennsylvania) and you host meetings recorded by Otter without explicit all-party consent, the litigation exposure may fall on you.
  1. The migration options are narrower than they looked. Revised 3 August 2026. ZoomMate is the best alternative for teams already on Zoom. Corrected 5 September 2026: this line previously said ZoomMate is included with paid Zoom plans. Zoom also lists a limited free ZoomMate Basic with Workplace Basic, and the fuller product is a separate licence carrying AI credits, so “included with paid plans” was too simple. Check the current plan detail before you migrate on the strength of it. Fathom has comparable accuracy and cleaner integration defaults than Otter, but trains its own models on de-identified meeting content by default — turn that off. Granola was previously named here as the structurally best alternative; that recommendation is withdrawn following the class action filed against it on 30 July 2026. Migration is still prudent, but it is a move between tools that share the same default, not an escape from it.

Sources

  • Otter.ai privacy and security documentation (otter.ai/privacy, verified 2026-05-24)
  • Court filings: In re Otter.AI Privacy Litigation, No. 5:25-cv-06911 (N.D. Cal.) and consolidated cases
  • Court filings: Cruz v. Fireflies.AI Corp., No. 3:25-cv-03399-SEM-DJQ (C.D. Ill.)
  • Mason LLP: Your AI Meeting Assistant May Be Stealing Your Voiceprint (April 2026)
  • UC Today: Otter.ai on Trial, and the AI Notetaker Industry with it (April 2026)
  • Babst Calland: Eavesdropping by Algorithm legal risks analysis (December 2025/January 2026)
  • Fisher Phillips LLP: New Lawsuit Highlights Concerns About AI Notetakers (August 2025)
  • Best Law Firms: When AI Listens Too Closely future of compliance (November 2025)
  • TLDV.io: AI Meeting Recorder Lawsuits 2026 comprehensive guide (May 2026)
  • HR Executive: A lawsuit over AI notetakers should be on every HR leader's radar (April 2026)
  • Jackson Lewis P.C.: AI Notetaking Tools Under Fire (August 2025)
  • AI CERTs News: AI Security Spotlight Otter.ai Zoom Recording Lawsuit (February 2026)
  • TechCrunch interview with Sam Liang (October 7, 2025)
  • Nature Communications: voice embedding speaker identity reconstruction study (2023)

Related on AI Leakage