Research: The Hidden Penalty of Using AI at Work
Artificial intelligence is rapidly transforming the workplace, but new research published in Harvard Business Review uncovers a hidden challenge: employees may face a social penalty for using AI tools, even when those tools improve efficiency. In a yearlong study of engineers using an advanced AI coding assistant, only 41% had adopted the tool. The numbers were even lower for women (31%) and employees over 40 (39%), showing that adoption is uneven across demographics.
To explore why, researchers ran an experiment with 1,026 engineers who were asked to evaluate Python code. The code samples were identical, but reviewers were told that some had been written with AI support. Even though the quality was the same, work labeled as “AI-assisted” received competence ratings 9% lower on average. This bias was most pronounced when evaluating work attributed to women or older engineers, suggesting that AI use may compound existing workplace inequalities.
These findings highlight that the biggest barrier to AI adoption may not be technical but social. Companies often focus on infrastructure, access, and training, assuming that once employees have the tools, they will use them. But employees make rational decisions based on how they believe their work will be perceived. If using AI makes them seem less competent (or reinforces existing stereotypes) they are understandably reluctant to adopt it.
For organizations, this means successful AI integration requires more than just technology. Leaders must address the cultural dynamics that shape employee behavior. That could mean normalizing AI use, rewarding transparency, and reframing AI as a tool for augmentation rather than a shortcut that undermines skill. Creating psychological safety around AI use is essential if organizations want equitable adoption across teams.
Ultimately, the research underscores that building an AI-enabled workplace is not just a technical challenge but a human one. Without tackling the stigma and bias associated with AI use, companies risk reinforcing inequalities and slowing down adoption. The message for leaders is clear: AI success depends not only on rolling out new tools but also on reshaping workplace culture to ensure all employees can benefit from them.