The noteakers are prepared to begin the assembly, however will they all the time be so welcome? A category motion lawsuit focusing on probably the most extensively used AI transcription instruments is drawing new consideration to a compliance hole for HR groups to fret about: the authorized dangers of AI-powered assembly notetakers within the office.
In re Otter.AI Privateness Litigation, now a consolidated case earlier than Choose Eumi Ok. Lee within the U.S. District Court docket for the Northern District of California, alleges that Otter.ai’s notetaking instruments recorded non-public conversations with out the consent of all contributors and used these recordings to coach its AI fashions with out ample disclosure. No substantive rulings have been issued but, however employment attorneys say the case is already signaling the place legal responsibility may land for employers.
“The AI transcription and recording concern is a scorching concern,” says Bradford Kelley, a shareholder at Littler Mendelson who co-authored a February 2026 evaluation of the litigation. He advised HR Government that human useful resource groups needs to be “very on this case.”
Kelley says that his agency will get fairly a number of questions from employers working in states which have all-party consent: “What do we have to do to ensure we’re in keeping with finest practices?”
In keeping with Littler, federal wiretap regulation and most state counterparts observe a one‑get together consent rule. Nonetheless, roughly a dozen states require all contributors to consent to the interception or recording of a dialog. A single digital assembly that features workers, clients or candidates in a number of jurisdictions can subsequently set off overlapping and generally inconsistent consent obligations that many employers haven’t totally mapped, in response to Littler.
Below the federal Wiretap Act, Littler factors out that non-public plaintiffs might search statutory damages calculated because the larger of a per‑day quantity or a minimal statutory award. On the identical time, Illinois’ Biometric Info Privateness Act (BIPA) authorizes statutory damages for improper assortment or use of biometric identifiers, together with when AI observe‑taking instruments determine particular person audio system by their voiceprints.
Learn extra: Congressional witnesses cut up on AI regulation, state legal guidelines stumble
Threat publicity areas
Kelley and co-author Zoe Argento, additionally a Littler shareholder, define seven threat areas employers ought to consider: consent, biometrics, accuracy, discrimination and disparate affect, attorney-client privilege, knowledge retention and confidentiality. The breadth of that record displays simply what number of authorized frameworks a single AI notetaker can activate directly.
On the discrimination entrance, the attorneys flag that AI transcription instruments might constantly misunderstand accents, speech impediments or different traits tied to protected courses. This could create disparate affect publicity if these transcripts inform efficiency opinions, hiring selections or disciplinary actions. Employers utilizing these instruments in employment decision-making can also set off AI-specific discover and audit necessities in jurisdictions together with New York Metropolis, Illinois and California.
Compliance for multinational employers
For multinational employers, the compliance image grows considerably extra advanced. Below the GDPR, the consent commonplace is a distinct and extra demanding consent framework than typical U.S. consent guidelines for name recording. Legitimate consent have to be freely given, particular and unambiguous from every particular person whose knowledge is processed. This implies a mannequin that depends on one assembly participant to authorize recording on behalf of all others would doubtless not fulfill the laws.
Information switch is a compounding concern, since recordings processed by U.S.-based distributors should adjust to worldwide switch mechanisms equivalent to Commonplace Contractual Clauses. And starting in August 2026, the EU AI Act introduces a separate layer of obligation. AI techniques used for employee monitoring and administration could also be categorized as high-risk, a class that might embody instruments providing sentiment analytics or productiveness scoring alongside transcription.
In co-determination international locations equivalent to Germany and France, deploying an AI notetaker can also require works council session earlier than rollout, a requirement with no U.S. equal that multinational HR groups often overlook.
What HR leaders can do about AI notetakers
The Littler evaluation raises a sensible actuality that HR leaders might discover uncomfortable: Banning AI notetakers outright is probably going unenforceable. One in 5 professionals reported often utilizing AI to draft assembly notes in a 2025 survey, and workers are bringing these instruments in whether or not or not employers have addressed them.
The attorneys’ suggestion is to get forward of it. Choose, configure and management a vetted software reasonably than cede that floor to no matter workers occur to obtain.
Meaning vetting distributors on knowledge safety and configuration choices, turning off options like voice recognition the place biometric threat is excessive, organising consent notices earlier than conferences, establishing brief knowledge retention durations and constructing clear insurance policies on when and the place AI notetakers are permitted.
The Otter.ai litigation has not but produced rulings that set binding precedent. However Kelley says HR leaders needs to be paying consideration now earlier than courts outline the boundaries for them.
