Employers are facing a new workplace hazard: AI notetakers that don’t know when to stop listening. In some virtual meetings, employees drop off the call while an AI assistant stays behind, quietly documenting gossip or disparaging remarks made by remaining employees, then emailing the transcript to the full team. 

“Those issues create some of the most excruciating problems,” says Joe Lazzarotti, an attorney at Jackson Lewis who is increasingly advising companies on AI notetaker mishaps. “One of the biggest issues is just trying to figure out how do we control that? How do we educate employees about it? And how do we make sure we’re not falling into a trap of the unwary?”

AI notetakers don’t always spell bad news. A new study from AI assistant startup Read AI in partnership with organizational behavior expert Rebecca Hinds found that recorded meetings lead individual contributors to speak nearly as much as managers, with women participating 9% more than men.

Still, the recording and dissemination of sensitive conversations creates real legal and HR risks. Lazzarotti advises leaders to conduct a full AI audit before problems surface, including reviewing governance, risk, and compliance rules; understanding what data is captured; deciding which meetings should be recorded; and controlling where recordings and documents are stored and who can access them.

Companies can also set clear rules around when AI notetakers are appropriate, require explicit disclosure and consent before recording begins, and limit how transcripts are distributed, avoiding automatic emails to entire teams. Some organizations are opting for summaries instead of verbatim transcripts or giving meeting hosts a “kill switch” to stop recording if conversations drift into sensitive territory.

Employee education is critical, too. Workers should understand that recorded meetings function more like written documents than private conversations and that offhand comments made after others leave a call may still be captured.

“[AI notetakers] could be a very powerful tool, but it’s just a matter of thinking it through before you go out and deploy it,” says Lazzarotti.

Editor’s note: This newsletter will be off for Presidents Day. Catch us back in your inboxes on Feb. 23.

Kristin Stoller
Editorial Director, Fortune Live Media
kristin.stoller@fortune.com

Around the Table

A round-up of the most important HR headlines.

More managers are using AI agents to help decide raises, promotions, and layoffs. Washington Post

Extra vacation time now beats bonuses as the top driver of feeling valued at work, according to new research. Wall Street Journal

“Storytellers” are emerging as one of tech’s hottest jobs, with salaries above $400,000. Business Insider

Watercooler

Everything you need to know from Fortune.

ROI debate. Obsessing over short-term ROI from AI tools is stifling innovation, and leaders need to loosen control, says Nvidia’s CEO. —Nick Lichtenberg

Old-school approach. A workspace CEO says bosses who cling to rigid, five-day mandates aren’t prepared for an AI-driven future. —Orianna Rosa Royle

New guard rising. With the average global CEO tenure down to 7.2 years, wider AI adoption could further reshape who makes it to the C-suite. —Phil Wahba