No AI Notetakers in the Boardroom, Please!

Why Your Board Shouldn’t Use AI Transcription Tools

Photo by Igor Omilaev on Unsplash

A few years ago, I joined a board meeting and noticed an unfamiliar participant listed: Otter.

I assumed it was someone’s assistant with an odd name… but then came the post-meeting minutes: A complete transcript of everything said, including a personnel discussion that was confidential. Otter turned out to be an AI notetaker, invited by our board secretary to ‘help’, that had recorded and distributed our entire meeting.

I was shocked. And, if I’m honest, a little embarrassed that I hadn’t noticed.

This experience isn’t unusual. Across boardrooms today, AI notetakers are showing up uninvited (or at least unannounced). An unfamiliar name appears in the participant list. A transcript lands in your inbox after a meeting you didn’t know was being recorded. A colleague mentions, almost in passing, that they ‘use AI for all their notes now.’ Perhaps, even, you realize the notetaker at the table is someone’s phone, propped up discreetly.

These aren’t just awkward moments.

As a board member, you are given access to sensitive information, financial details, and strategy that are confidential by necessity. When AI enters that room, the consequences can extend well beyond embarrassment…. They reach into legal exposure, eroded boardroom trust, and a compromised ability to govern.

Here’s what every board member should understand.

---

Jump to a section:

• ⚠️The Real Risks

• 🚨The problem is just getting worse.

• 🙋What You Can Do As A Board Member

• 🎯A Final Note

• 👉Next up in my AI & Governance series…

---

⚠️The Real Risks

1. Everything You Say Suddenly Becomes a Legal Record

Board meeting minutes should be carefully curated, capturing key decisions, discussion points, and rationale. What they include is as important as what they leave out. Preliminary thinking, exploratory debate, and half-formed ideas are part of how good boards work. They are not meant to be preserved verbatim.

AI, on the other hand, transcribes everything.

That offhand comment made while still settling into the room. The frustrated remark before a difficult vote. The joke a colleague made to break tension. None of it is filtered and all of it is likely discoverable in litigation.

Consider a director who arrives flustered and jokes: ‘I barely had time to glance at the materials this week.’ Everyone laughs and moves on. But if that board later faces a contested decision, that offhand comment is now a record of a director who, by their own words, voted without preparation. That's a breach of the legal duty of care waiting to be argued.

Read more about the legal risks + here + here too!

2. Your Confidential Information Doesn't Necessarily Stay in the Room

When AI records your meeting, that conversation leaves the room immediately. It moves to vendor servers, where it may be stored indefinitely under data retention policies you didn’t negotiate and can’t control. Vendor employees may have access and the content of your discussions may be used to improve the AI product itself.

Think carefully about what that means: Personnel discussions, financial commentary, strategic considerations…. All of it now held by a third party, under their terms, not yours.

The legal consequences reach further. Attorney-client privilege, one of the most important protections available to a board navigating serious risk, can be waived when confidential communications are disclosed to a third party. If your board engages legal counsel in confidence and an AI tool is routing that conversation through an outside vendor, that privilege may already be forfeit.

3. AI Dampens Candor

Good governance depends on frank conversation. Directors need to feel free to debate, challenge assumptions, float half-formed ideas, and voice minority perspectives without worrying how it will look on the record.

When sensitive topics are on the agenda and AI joins the room, you can bet that people won’t be as forthcoming. I’ve seen board members who are usually more direct chose careful words over candid ones, and some not participate at all. And their caution isn’t unfounded. AI misses context, flattens tone, and, when multiple people speak at once, mix up words and makes mistakes.

So it’s not so safe to speak up.

But a board that can’t speak up can’t govern well.

4. You May Be Breaking the Law

Federal law requires only one-party consent to record a conversation. But eleven states, including California, Florida, Illinois, Pennsylvania, and Washington, require all parties to consent. If even one participant is located in one of those states, that standard applies to everyone in the meeting.

Directors joining late, missing a consent notice, or dialing in from a stricter jurisdiction create legal exposure, for themselves and for the organization. It doesn’t require bad intent. It just requires not knowing, and that is increasingly easy in a world where recording features are being built directly into the platforms your board already uses.

One More Thing: It’s Not Even More Efficient

The pitch is that AI notetakers save time. The reality is that someone still has to convert an hour-long verbatim transcript into proper minutes.. and an unfiltered record, full of errors and crosstalk, is harder to work from than a well-prepared set of notes.

👀Unless, of course, you're using a second AI to summarize the first one. Which raises its own questions and opens an entirely new layer of risk.

The boardroom is one of the few places where the promise of efficiency should not be the deciding factor.

---

🚨The problem is just getting worse.

Unfortunately, AI transcription is becoming unavoidable. Zoom, Microsoft Teams, Gmail, and other platforms are building these features directly into their products … and this isn’t always something you choose. Maybe it gets enabled by default when you update your software, or turned on by a single participant without announcement, or activated accidentally.

There also is a less-named force at work: tech companies really want you to use it. There’s been a flood of content being released praising how AI can ‘transform board governance’. It reads authoritative.. but often these are marketing materials produced by companies with a financial interest in your adoption or individuals getting a kick-back. Their enthusiasm isn’t neutral

The reality:?

⭐AI transcription in board meetings presents well-documented risks that can compromise confidentiality and decision-making integrity.

---

🙋What You Can Do As A Board Member

Watch for these red flags:

• 🚩Unfamiliar names in the participant list.

• 🚩Automated ‘this meeting is being recorded’ messages you didn’t authorize.

• 🚩Transcript emails after meetings.

• 🚩Colleagues treating AI note-taking as routine/expected.

• 🚩No board policy on recording or AI tools.

The first line of defense is simply paying attention. Before and during any meeting, watch for unfamiliar names in the participant list. Tools like Otter, Fireflies, and Notetaker are common, but they don’t always announce themselves clearly. An automated ‘this meeting is being recorded’ message you didn’t authorize is another signal. So is a transcript arriving in your inbox after a meeting you thought was private.

If a colleague mentions they use AI for all their notes now, that conversation is worth having directly. What tool? Where does it store data? Who has access? These are reasonable questions, not paranoid ones.

It’s OK to say no.

If you join a meeting and see AI at work, you have the right to raise the question. For most private organizations, even in one-party consent states, board members typically have a reasonable expectation of privacy, especially in executive session.

If recording continues over your objection, document it and leave if necessary. You’re not being difficult…. you’re protecting the organization and yourself.

(Note: boards of public institutions operate under different standards and should seek specific legal guidance.)

Advocate to get a policy in place.

Your board needs a written policy that addresses what may be recorded, which tools are approved for use, and where AI fits into your governance processes. This isn't optional. It's governance hygiene for the modern era.

And the boardroom is just the starting point. AI is quietly making its way into every corner of organizational operations: communications, fundraising, HR, finance. Your policies need to grow to meet that reality. But it starts here, with the room where your most consequential decisions are made.

---

🎯A Final Note

I say this as someone who grew up in a tech household and has always been excited about what new tools can do. Technology amplifies human choices, for better or worse.

In nonprofit governance, where trust and honest deliberation are foundational, we need to be thoughtful about which tools we invite into the room. AI transcription may work well in other lower-stakes contexts….The boardroom is not one of them.

---

Keeping Reading Part II >

So Your Board Said Yes to AI. Here's How Not to Regret It.

---

💡P.S.! If your board insists on using AI despite these risks…

Subscribe for free and follow along. I’ll be covering the practical steps legal experts say boards must take to reduce exposure, including how to begin building your AI policy, and how to address the broader wave of AI tools quietly making their way into every platform your organization already uses.

---

🔗References & Resources

• https://www.shumaker.com/insight/ai-notetakers-in-the-boardroom-privilege-and-privacy-considerations/

• https://corpgov.law.harvard.edu/2025/11/29/using-ai-in-the-boardroom-new-opportunities-and-challenges/

• https://www.bsk.com/news-events-videos/artificial-intelligence-in-the-boardroom-key-legal-governance-amp-risk-considerations

• https://www.wagenmakerlaw.com/blog/always-listening-ai-automatic-note-takers-and-related-risks

• https://www.babstcalland.com/news-article/eavesdropping-by-algorithm-legal-risks-of-ai-meeting-assistants/