4 AI Etiquette Rules That Actually Matter at Work Right Now

What the data says

Two recent surveys paint a clear picture of how AI is showing up in workplace communication, whether anyone asked for it or not. Thirty-seven percent of employees say they have used AI tools for tasks that require emotional intelligence, according to a May 2026 survey from GoTo and Workplace Intelligence. Nearly as many, 35 percent, told ZeroBounce in September 2025 that they had used AI to draft or edit sensitive workplace communication. At the same time, researchers from the University of Kent found that people who outsource socio-relational tasks to AI are perceived more negatively by their peers. The numbers point in one direction: AI use at work is widespread, but using it for the human parts of a job comes with a social cost that most people have not caught up to yet.

The four rules experts keep coming back to

CNBC Make It spoke with several workplace experts this week about where AI etiquette is right now, and four rules surfaced repeatedly. First, ask before adding an AI notetaker to a meeting. It has become common practice, but Lindsey Pollak, a career and workplace expert, says an unannounced AI participant makes the interaction feel like an interrogation, not a business call. Second, do not outsource genuinely human tasks to AI. Tessa West, a social psychologist at NYU, points to research showing that managers who use AI for interpersonal messages are seen as less authentic and trustworthy. Apologies, expressions of gratitude, and anything meant to convey sincerity are the worst candidates for AI assistance. Third, do not cite AI as your source of advice in a workplace conversation. Opening a salary negotiation with 'I asked ChatGPT what I should ask for' signals that you value the chatbot's input more than your manager's experience. Fourth, agree on shared norms. West recommends workplaces develop explicit rules about when AI use is appropriate and when it is not. Ambiguity breeds suspicion, and suspicion erodes trust faster than any misplaced comma.

The honest nuance

None of the experts are telling people to stop using AI at work. Jacqueline Whitmore, founder of the Protocol School of Palm Beach, advises editing AI output and adding a personal touch so it does not read as a canned response. You do not need to lose your voice. You just need to make sure your voice is the one coming through. West's point about shared norms is also a reminder that the problem is not AI itself. It is the ambiguity that AI introduces into interactions that used to be straightforward. When someone sends a thoughtful message, you used to know they wrote it. Now you do not know, and that uncertainty alone changes the dynamic. The fix is not less AI. It is more clarity about when and how you use it.

What to actually do

Start with one small, visible norm-setting move this week. If you use an AI notetaker in meetings, add a one-sentence ask at the top of your next invite: 'I'd like to use an AI notetaker so I can stay present in the conversation. Let me know if that works for you.' If you use AI to draft emails or messages, add one personal sentence before sending. Not a rewrite of the whole thing. One line that could only have come from you. A reference to something specific you discussed last time, a genuine reaction, a short update. The goal is not to hide that you use AI. It is to make it clear that the person on the other end is still talking to a person.

From Kindloom Labs

If you want a plain-English guide to what AI is actually good at and what it isn't, the free Quick Read What AI Can and Can't Do covers the fundamentals in about four minutes. Knowing where the line is between a tool and a substitute makes these etiquette decisions a lot easier.

Sources

• To avoid sounding 'artificial and rude,' follow these AI etiquette rules, experts say (CNBC Make It, Jul 15 2026)

Keep reading