The AI Notetaker Is Sitting in Your Meeting Uninvited

The AI Notetaker has quietly become a coworker. Roughly a third of employed Americans say an AI notetaker or transcription bot — tools like Otter.ai, Fireflies, Zoom AI Companion or Microsoft Copilot — has been present in their work meetings, according to a July 2026 survey of 500 employed U.S. adults commissioned by Kolmogorov Law and conducted through the Pollfish research platform. More than a fifth could not even say whether they had been recorded or not, which is its own kind of finding.

What has not become routine is asking. Among the workers whose meetings included an AI Notetaker, only about a third said they were always asked for permission before the tool recorded or transcribed. More than a third were asked only sometimes, a quarter were never asked and simply watched the bot appear in the call, and a small remainder learned about the recording only afterward. Put together, nearly two-thirds of exposed workers said consent was inconsistent or absent. The consent problem is not limited to bots either: close to one in five of all workers surveyed said they had discovered after the fact that a meeting had been recorded or transcribed without their knowledge, and a smaller share said it had happened more than once.

The issue is already reaching the courts. A federal class action against one major transcription app alleges that meetings were recorded and transcribed without the consent of all participants in violation of California privacy law, the survey report notes. The company denies the allegations, and the case remains pending. Meanwhile, an Atlassian employee's attempt to use AI-generated meeting notes as legal evidence collapsed this month: Australia's Federal Court ruled on October 13, 2026 that the "transcript" his employer produced was not a transcript at all but an AI summary, and declined to treat it as proof of a workplace agreement, according to HRD reporting.

Your words are becoming a business record

Employment lawyers say the real risk is not just rudeness — it is permanence. AI notetakers and meeting assistants that transcribe, analyze tone and flag inconsistencies create business records that are discoverable in litigation, and those transcripts are frequently inaccurate because nobody goes back to correct them, according to a workplace briefing by FRB Law. As the briefing put it, anything typed into an open AI tool should be treated as fully discoverable: "never write or paste anything into a chatbot that you wouldn't want read back to you in a deposition."

The recording question also sits on shaky legal ground. Recording laws vary sharply across jurisdictions — one-party consent states like New York sit next to two-party consent states like California — but lawyers note the practical reality is that employees are recording and transcribing conversations constantly, regardless of what the law technically requires. Add "shadow AI" to the mix: workers bypassing secure, enterprise-grade tools in favor of personal accounts, often dropping confidential or privileged information into platforms with no confidentiality protections at all.

The technology is also moving faster than workplace norms. Amazon introduced two AI wearable devices, the Bee Frontier and Bee Eclipse, on October 15, 2026, designed to turn everyday conversations into transcripts, summaries and to-do lists. A former senior security engineer at the company argues the devices' consent safeguards put the responsibility on the wearer rather than giving other participants a practical way to refuse being recorded, according to TechRepublic. The concern is straightforward: what happens when an employee records a meeting without everyone else's permission?

What you can actually do about it

First, know that asking is normal. "Respectful pushback is not unprofessional — it's prudent," employment attorneys at Coblentz Law advise. If you join a meeting that is being recorded and you did not consent, you have every right to speak up, ask for the tool to be disabled, or leave the meeting if needed. At a minimum, request a copy of the transcript afterward and review it for accuracy.

Second, treat every transcribed meeting as written. Speak as though your words could be replayed in a review, a dispute or a deposition — because legally, they might be. Save the candid brainstorming for channels where nothing is being recorded, and keep anything sensitive out of AI-assisted calls entirely.

Third, ask your employer what the policy is. Lawyers recommend companies be intentional: don't record by default, get explicit consent, vet vendors, limit how long transcripts are kept, and update handbooks to say when AI notetaking is appropriate. If your workplace has none of that, the survey data suggests you are in the majority — and raising the question is the first step toward a real answer.

The productivity promise of these tools is genuine — for workers who attend several calls a day, auto-generated summaries and searchable transcripts are a real time saver. But the survey shows the bargain is being made without the people paying the price. Until asking becomes as routine as recording, the smartest productivity move in any meeting might be the oldest one: saying "not today" to the bot.

For more on how AI is reshaping the workday, see The AI Productivity Perception Gap Is Wider Than You Think and The AI Productivity Paradox: Workdays Keep Getting Longer, plus OpenAI Models on Google Cloud: Enterprise AI's New Normal on the enterprise AI push driving these tools into offices.

Sources: the consent survey via Kolmogorov Law; workplace consent analysis via TechRepublic; the AI-notes court ruling via HRD Australia; employer guidance via Coblentz Law.