'Everybody and their mother is using these things': AI note-takers are creating a legal time bomb in boardrooms — and most executives have no idea

But corporate lawyers are increasingly warning that those same tools could create a legal nightmare for companies that don't understand the risks.

"Everybody and their mother is using these things," San Antonio attorney Jeffrey Gifford told The New York Times (1). "Executives are using them, boards are using them, non-executive businesspeople are using them."

The problem, lawyers say, is that AI note-takers don't just capture the important parts of meetings. They capture everything: offhand comments, jokes, speculation, corrected statements and casual remarks that previously might never have been written down at all.

And once those records exist, they can potentially become evidence or fall under the scrutiny of regulators.

Why lawyers are suddenly nervous

Tools that automatically record and summarize meetings have grown alongside the rise of workplace AI. Many video conferencing platforms now offer built-in transcription tools, and standalone AI assistants promise detailed summaries, task assignments to meeting participants and searchable archives of conversations.

Their appeal is obvious: fewer missed details and less administrative work, but lawyers argue the convenience comes with hidden consequences.

In lawsuits or government investigations, companies are often required to turn over documents and communications related to disputed issues. Historically, formal board minutes or meeting notes were carefully curated and limited in scope.

AI-generated transcripts change that dynamic. Offhand remarks — like a joke about a rival or a potential deal — can be preserved and later scrutinized by regulators or a judge. Even a passing comment about a known risk might resurface years down the line as evidence in a shareholder lawsuit.

"Typically, private litigation asks for all documents and communications related to a particular topic, so it's not so much that a litigant needs to specifically ask for it," Christoffer Lee, a white-collar defense lawyer, told the Times, adding that he expects regulators to begin seeking these transcriptions in their scrutiny of companies.

The bigger fear: losing attorney-client privilege

For many companies, the greatest concern involves attorney-client privilege, the legal protection that generally keeps conversations between lawyers and clients confidential.

Some attorneys worry AI note-taking tools could unintentionally weaken or even waive that protection. After all, if sensitive legal discussions are shared with outside AI companies through transcripts or recordings, courts could potentially view that as disclosure to a third party (2).

Courts are only beginning to wrestle with the issue. Earlier this year, a judge ruled that legal advice generated through the AI chatbot Claude was not protected by attorney-client privilege because users had limited expectations of privacy under the platform's terms (3).

Another federal judge reached a different conclusion involving ChatGPT usage, highlighting just how unsettled the legal landscape remains (4).

Why this matters beyond boardrooms

The issue isn't limited to Fortune 500 companies.

Small businesses, startups and even regular office workers increasingly rely on AI meeting assistants without fully understanding where recordings are stored, who can access them or how long they remain available.

Many users also assume AI-generated summaries are perfectly accurate. Lawyers warn that's dangerous: an incorrectly transcribed sentence — especially involving legal, financial or compliance matters — could later become difficult to challenge if participants no longer remember the original conversation clearly.