Most nonprofit leaders have had the same awkward Zoom moment: you join a meeting, someone says hello, and then a second “someone” appears with a name like “Otter.” At first, it seems great: no one has to take minutes, remember who promised to follow up with the finance committee, or decode handwritten notes.
But AI note-takers are, fundamentally, recording and transcription software branded as convenience tools, and for nonprofits, where ordinary meetings can include everything from sensitive board strategy to HR concerns, it’s critical to consider where meeting records end up.
AI note-takers can create records you didn’t mean to create
Record creation should be top of mind when a note-taker enters the room; a meeting that used to result in a short summary might now produce a transcript or recording (that might not even be accurate). That’s a different kind of record, and it could become evidence down the road. If the organization is later involved in litigation (like a donor or employment dispute) or a government inquiry, those AI-generated notes might be discoverable, and a casual comment might live in a searchable file with a timestamp and speaker labels.
Nonprofits also already have official records for certain meetings, especially board and committee meetings. Board minutes generally capture actions taken, key decisions, recusals, conflicts, and materials reviewed. They are not supposed to be a word-for-word transcript of everyone’s stray thoughts. While AI notes can be helpful if they are accurate and properly retained, when they are misleading or inconsistent with the organization’s official records, they can create an unhelpful paper trail.
The takeaway here for nonprofit leaders is to carefully consider which meetings your organization actually wants that kind of record for and not to allow unintentional creation of records to proliferate.
Some nonprofit meetings should be bot-free
AI note-takers don’t belong in every meeting just because they effortlessly create tidy bullet points. Sensitive meetings like board executive sessions; conferences with legal counsel; employee and investigation discussions; medical, accommodation, or whistleblower conversations; and donor strategy meetings should be treated carefully.
These types of meetings raise unique concerns. For example, if an AI note-taker joins a meeting with legal counsel, the organization needs to think carefully about whether the tool itself, vendor access, storage settings, or later sharing could create privilege problems. In other contexts, sensitive data is a concern: a transcript might capture donor, employee health, or other information the organization did not intend to store in the AI’s system. And watch out for when your organization’s privacy policy says it collects, uses, retains, or shares information one way, but the AI meeting tool is doing something else.
Consent also matters. Recording rules vary by state, and an AI tool might trigger consent issues depending on how it records or shares meeting content.
A good policy generally will make clear that sensitive meetings require approval before an AI tool is permitted to attend.
Nonprofits should approve AI tools before staff use them
Nonprofits should decide which AI note-taking tools are approved before staff start inviting them to meetings. The organization should know where approved tools store data, who can access transcripts, whether the vendor uses meeting content to train models, how long recordings and transcripts are kept, whether and how the content can be deleted, and what happens if there is a legal hold.
While this sounds like a lot (and it is), creating an organizational policy doesn’t need to become an overwhelming project. Fundamentally, a comprehensive policy might indicate that staff may use only approved AI note-taking tools, require participant consent before recording or transcription and specify the types of meetings where the use of AI note-takers is banned (e.g., executive sessions, legal meetings, HR investigations, and other sensitive meetings unless legal or senior leadership approves an exception). It also could require users to delete transcripts when they are no longer needed, unless the organization has a legal, grant, governance, or legitimate operational reason to keep them.
Your AI, privacy, recording, and retention policies should agree with each other
Your AI policy, privacy policy, meeting recording policy, and document retention policy all should be compatible with one another when it comes to AI note-takers. If one policy says transcription always is allowed, another says meetings can’t be recorded without approval, and the privacy policy doesn’t contemplate this kind of data collection at all, your organization might have set itself up for confusion.
Most meeting recording policies should define “recording” broadly enough to include audio, video, transcripts, automated summaries, chat logs, and AI-generated meeting notes.
Meeting recording policies also generally cross-reference the organization’s document retention policy. If a document retention policy says routine meeting recordings or transcripts should be deleted after a certain period, the AI tool and Zoom or Teams settings should match that rule. That means that auto-delete settings should be reviewed before the tool is rolled out, and the organization should decide whether AI transcripts and notes are deleted automatically (and if so, who can override deletion), what happens if a legal hold is in place, and where “deleted” content actually goes and for how long.
And if the organization has different retention rules for board minutes, personnel records, donor files, legal matters, grant records, and routine staff meetings, the AI note-taking policy should not necessarily collapse notes into one default setting.
Before inviting the bot, ask these questions
While AI note-takers can be useful, be wary of letting your bots into every room without rules. Nonprofits should treat AI meeting tools like any other system that touches sensitive information.
If your organization already has started using AI note-takers, consider taking an inventory as a first step. Find out which tools staff are using, where transcripts are stored, and which meetings are being recorded and under what circumstances. Then compare your AI policy, privacy policy, meeting recording policy, and document retention policy.
Before inviting the bot to a meeting, ask a few basic questions: Are we allowed to record or transcribe this meeting? Does every participant know the AI tool is present and agree to its presence? Could this meeting involve legal advice, HR issues, donor strategy, client information, or other sensitive data? Does our privacy policy allow this kind of collection and storage? Does our retention policy tell us how long to keep or delete the transcript? Are the tool’s auto-delete settings actually set up that way? And if this transcript showed up in discovery later, would we be glad we had it?
__________________________
If your nonprofit is using AI note-takers, or if staff already have started bringing their AI “plus one” to meetings, now is a good time to check whether your policies have your organization covered.
I’ve put together a Nonprofit AI Note-Taker Policy Checklist to help organizations review AI tools, meeting recording rules, privacy language, retention schedules, auto-delete settings, and internal approval processes before the next mystery transcript appears.
You can download it here.
And if your organization needs help turning this checklist into actual policy language or aligning your privacy, recording, and document retention policies, reach out about working together 1:1.
