LiveSuggest
Back to blog

Do You Need Consent to Use AI in Meetings? (2026 Guide)

AI meeting tools are now mainstream: 75% of professionals use one in their work meetings, making them core workplace infrastructure (Source: Fellow.ai, 2025). But rapid adoption has surfaced a serious tension — most of these tools record your meetings, and recording conversations raises real legal and ethical questions.

In many jurisdictions, you can’t simply start recording a meeting without telling everyone. Some countries require consent from all participants. Others require only one-party consent. And the rules vary significantly between countries and even between US states.

For teams using AI tools across borders — which applies to most remote teams today — this creates a compliance challenge that’s hard to ignore. Understanding where the legal lines sit starts with understanding how these tools actually work.

The consent question depends almost entirely on one thing: does the tool create a recording?

Model 1: Bot-based tools (recording)

Traditional AI meeting assistants send a bot into your call. The bot records everything, processes the audio, and produces a summary after the meeting. This means:

Model 2: Real-time processing (no recording)

Some AI tools process audio in real time — capturing it from your microphone or browser tab, running it through AI processing, and discarding it immediately. No recording is created, no audio is stored. This means:

For a deeper look at how no-recording tools work, see our guide on no-bot meeting assistants.

The Privacy Paradox: Why This Matters for Meeting Culture

Consent isn’t just a legal box to check. Recording changes how people actually behave in meetings.

According to Fellow.ai’s 2025 survey of professionals, 84% of users say they modify what they say when an AI note-taker is present. The meeting gets captured — but the conversation itself becomes more guarded. Separately, 50% of professionals who haven’t adopted AI meeting tools cite privacy and security as their primary barrier to doing so.

For any team leader or organization deploying meeting AI, this behavioral impact is as consequential as the legal question. A tool that makes participants guarded undermines the very meetings it’s supposed to improve.

This dynamic is one of the strongest practical arguments for choosing tools that minimize their data footprint — not just for compliance, but for team trust.

GDPR: What Actually Matters

GDPR is the most relevant framework for European teams, but its principles influence privacy law worldwide. Here’s what matters for meeting AI:

Data minimization (Article 5(1)(c))

GDPR requires you to collect only the data you need, for only as long as you need it. A tool that records entire meetings and stores them is harder to justify under this principle than one that processes audio in real time and immediately discards it.

Legal basis for processing (Article 6)

Recording a meeting typically requires either explicit consent from all participants or a documented legitimate interest assessment. Real-time processing without storage is easier to justify — especially when the purpose is personal productivity, not surveillance of others.

Storage limitation (Article 5(1)(e))

Personal data may not be kept longer than necessary for the stated purpose. If no data is stored, this principle is satisfied by default — one of the strongest compliance arguments for no-recording tools. The ICO’s official guidance confirms that organizations must determine and justify their retention timeframes.

Data subject rights (Articles 15–22)

When recordings exist, participants have the right to access, correct, or delete their data. If no recording exists, these rights are simpler to manage — there’s nothing to access, correct, or delete.

Outside of GDPR, many jurisdictions use a “party consent” framework for recording conversations:

Consent typeWhat it meansExamples
One-partyOnly one person in the conversation needs to consent to the recordingMost US states, Netherlands, South Korea
Two-party (all-party)All participants must consent before recording beginsGermany, France, Spain, UK (business context), California, Illinois

Note on the UK: Under UK GDPR and the Data Protection Act 2018, recording in a professional context requires informing all participants. The one-party exception in UK law applies to personal use only, not workplace meetings.

Note on California and Illinois: The California Supreme Court confirmed in 2025 that all-party consent applies to call recordings, including those involving non-participants. Illinois imposes criminal penalties for violations.

With a no-recording tool, this entire distinction becomes largely irrelevant — there’s nothing to consent to, since no recording is created.

How No-Recording Tools Simplify Compliance

The practical impact of using a tool that doesn’t record:

  1. No consent collection needed for recording — you should still be transparent, but the legal burden is dramatically reduced
  2. No data retention policy required — nothing is stored, so there’s nothing to retain or delete
  3. No breach notification for recordings — if there’s no recording, it can’t be leaked or accessed improperly
  4. Simpler vendor assessment — your legal team has fewer concerns when no personal data leaves your device
  5. Cross-border simplicity — the varying consent laws between countries matter less when there’s no recording

Before You Choose: 5 Questions to Ask

When evaluating any AI meeting tool, ask these questions to understand your consent obligations:

  1. Does this tool record audio? If yes, consent requirements apply in most jurisdictions.
  2. Where is data stored? On-device only, or on the provider’s servers? Server storage creates additional GDPR obligations.
  3. How long is data retained? Indefinitely? 30 days? Or is it immediately discarded?
  4. Does a bot join the meeting? A visible bot may trigger participants to ask about recording and consent.
  5. What’s the provider’s data processing agreement? Even if the tool claims “no storage,” verify this in writing.

Where LiveSuggest Stands

LiveSuggest is built around the no-recording model. It processes meeting audio in real time — via your microphone or browser tab audio sharing — and discards the audio immediately after processing. No bot joins your call. No recording is created. No audio is stored on any server.

This architecture was chosen specifically to minimize consent obligations and make the tool usable in privacy-sensitive environments. For details on how the audio processing works, see our technical overview of how meeting AI works without a bot.

Frequently Asked Questions

Do I need to tell participants I’m using an AI meeting tool?

It depends on whether the tool records. If an AI bot joins and records your meeting, you generally need explicit consent from all participants. If the tool processes audio in real time without recording or storing anything, the consent requirements are significantly reduced — though transparency with colleagues is always recommended as a matter of professional ethics.

Under GDPR, the key factor is whether personal data is stored. Bot-based tools that record and store meeting audio create personal data subject to GDPR obligations. Real-time processing tools that immediately discard audio avoid most of these obligations, making them easier to use legally in Europe. However, always check your specific country’s laws, as some have requirements beyond GDPR.

What’s the difference between recording and real-time processing for GDPR?

Recording creates a stored copy of personal data (voices, statements), triggering full GDPR data protection obligations: storage limitations, right to deletion, breach notification, and more. Real-time processing analyzes audio as it streams and discards it immediately — no personal data is retained, so most GDPR storage obligations don’t apply.

Healthcare and legal professionals face stricter rules around recording (HIPAA, attorney-client privilege, professional secrecy laws). AI tools that don’t record or store any audio are generally more compatible with these requirements, since no patient or client data is retained. Always verify with your compliance team for your specific jurisdiction and use case.

Conclusion

The consent question around AI meeting tools isn’t as complicated as it seems — it mostly comes down to whether the tool records. If it does, you need consent. If it doesn’t, the equation changes dramatically.

As AI meeting tools become standard in professional settings, choosing one that minimizes your compliance burden isn’t just convenient — it’s responsible. A no-recording, real-time approach lets you benefit from AI assistance without creating legal exposure for yourself or discomfort for your colleagues.

This article is for informational purposes only and does not constitute legal advice. Consult a qualified legal professional for guidance specific to your jurisdiction.


Sources

  1. The State of AI Meeting Notetakers 2025 — Fellow.ai, 2025 — adoption rates, privacy concerns, and behavioral impact statistics
  2. Art. 5 GDPR — Principles relating to processing of personal data — gdpr-info.eu — data minimization (Article 5(1)(c)) and storage limitation (Article 5(1)(e))
  3. Art. 6 GDPR — Lawfulness of processing — gdpr-info.eu — legal basis for data processing
  4. Storage limitation principle — ICO (UK Information Commissioner’s Office) — official guidance on GDPR storage limitation
  5. California Supreme Court Requires All-Party Consent to Record Phone Calls — National Law Review — California all-party consent confirmation
  6. Netherlands Audio and Video Recording Laws — RecordingLaw.com — Netherlands one-party consent framework
  7. 57 Meeting Waste Statistics for 2026 — MeetingToll — unproductive meeting cost data