Workplace AI can dull your edge if it replaces your thinking, professors warn

Artificial intelligence is already part of daily work for many Americans, and experts say the career risk is not the tool itself. The problem starts when workers use AI to outsource the thinking, judgment and personal communication that help them stand out.

A March study from the Federal Reserve Bank of St. Louis found that 43% of U.S. workers report using AI on the job. Sandra Matz, a professor at Columbia Business School, told CNBC that AI can make workers less distinctive when they rely on it for critical thinking or creative reasoning.

Matz said many AI systems tend to produce responses that are safe, familiar and widely shared. If employees lean on those outputs too heavily, she said, they risk becoming complacent and less original over time.

AI can help with busywork, but experts warn against handing over judgment

Several workplace researchers told CNBC that AI can be useful when it handles repetitive or time-consuming tasks. Matz said the technology can improve productivity by speeding up routine work, and it can serve as a brainstorming partner.

Americus Reed II, a marketing professor at the Wharton School of Business, said AI is better suited for summarizing information, reviewing data and helping organize ideas. Reed told CNBC the goal should be to reduce the drag of processing information while leaving judgment and meaning-making to the worker.

Fred Oswald, an industrial-organizational psychologist and former member of the National Artificial Intelligence Advisory Committee, said employees and companies are still learning where AI helps and where it adds risk. He described current workplace use as a fast-developing area that needs clearer understanding.

Dorothy Leidner, a business and ethics professor at the University of Virginia’s McIntire School of Commerce, said newer workers may face a higher risk. According to Leidner, employees without deep expertise may have a harder time using AI as a tool that strengthens their thinking instead of one that generates the thinking for them.

AI-written messages may weaken workplace trust

Matz also warned about using AI to write routine messages to colleagues, managers or employees. She told CNBC that when people suspect a note has been produced by AI, they may question whether the sender cared enough to write it personally.

AI can still be useful for summaries, meeting notes or status reports, Matz said. Her concern is informal workplace communication, where tone, spontaneity and even imperfection can signal attention and effort.

Leidner raised a related concern about “digital twins,” meaning AI versions of a person that can communicate on their behalf. She said if AI writes emails and AI answers them, real communication between people can disappear.

Identity at work is part of the challenge

Eva Selenko, a work psychology professor at Loughborough University in England, told CNBC she is especially concerned when workers use AI for tasks tied closely to how they see themselves as professionals. If a tool performs well at a core part of someone’s expertise, she said, it can feel like a threat to status and identity.

Reed said that reaction is understandable as AI takes on work people once considered central to their roles. He told CNBC that workers should connect their value to qualities AI cannot replicate in the same way, including judgment, creativity, curiosity and relationships.

The common thread from the experts is direct: AI may be useful at work, but relying on it too heavily can weaken the skills and human signals that careers are built on.

This story draws on original reporting from CNBC.