Workplace AI Favorability Cut in Half Over 5 Years

Negative perceptions of AI are spreading among employees who use AI tools in the workplace. An analysis of reviews posted on Glassdoor, a U.S. job and company review site, reveals that the proportion of positive mentions regarding AI dropped from 81% in 2019 to 43% over approximately five years, nearly cutting in half.

To understand this shift, it is necessary to review how AI has become integrated into the workplace. Around 2019, AI adoption in the workplace was relatively limited, and for many employees, AI was perceived with anticipation as "something new that will come." However, subsequently, particularly following the rise of large language models (LLMs), the implementation of AI tools in business operations accelerated rapidly. The initial anticipation for new technology gave way to real work experience, and attitudes shifted accordingly.

The dissatisfaction revealed in this analysis extends beyond concerns about job loss. Employees reported complaints such as "being forced to use AI," "work monitoring through AI is intensifying," and "being assigned unrealistic productivity targets due to AI." These voices are characterized more as criticism of how AI is being implemented and used in the workplace, rather than discontent with the AI tools themselves.

Moreover, it has become clear that evaluations of AI differ significantly depending on one's position. While executives generally view AI positively, negative assessments nearly dominate at ground-level positions such as insurance claims processors. Within the same organization, there is a clear temperature gap between the "side using AI" and the "side being required to use AI."

This temperature gap is not unrelated to the structure in which AI adoption discussions have traditionally centered on management perspectives. While focusing on management metrics such as productivity improvements and cost reduction, the actual daily experiences of frontline employees tend to remain invisible. The analysis of Glassdoor reviews—aggregating "voices from the field"—has effectively made this gap visible.

As AI becomes established in the workplace, this survey finding carries significant implications for companies. Rather than evaluating tools solely on performance and efficiency, there is a need to continuously understand how employees actually feel and to reconsider the implementation approach itself. Forced adoption or unilateral monitoring increases can actually harm overall organizational productivity and employee retention rates.

The key point to watch going forward is how companies address not only the "results" of AI adoption but also the "process." If employee dissatisfaction continues to accumulate, organizations risk undermining trust within the company rather than maximizing the benefits of AI adoption. How companies incorporate employee voices into policy and implementation strategy will be one of the decisive factors in the success or failure of corporate AI strategy.