Why do we fear robots at work? What does their presence actually trigger in us?
When companies introduce a robot to the floor, they often focus on the technical side, the part that's easy to explain and implement, because tech does what it's asked to do. What gets missed is the subtler part: what people think, and how they'll react to it. They worry about how the robot looks. Too mechanical and people keep their distance. Too human and people get unsettled, the uncanny valley problem we've talked about before in this newsletter. So companies spend real budget softening the robot's face, its voice, its gestures, on the theory that a friendlier-looking machine will be an easier coworker to accept.
A new study says that budget is solving the wrong problem.
Xian Zhao, now at Ohio University, ran five experiments with Geoffrey Leonardelli at the University of Toronto's Rotman School and Xiaoxiao Zhang at Shenzhen University, across more than 2,500 participants in China and Canada. The setup was simple: tell someone they scored lower than a robot on a work task, then measure what happens to how they feel about their job. Across every version of the experiment, the pattern held. People who believed a robot had outperformed them reported feeling less secure in their role, less committed to the company, less satisfied at work.
Here's the part that should worry anyone who budgeted for a friendlier-looking robot: making the robot more humanlike changed nothing. The threat people felt had nothing to do with whether the robot had eyes or a name or a warm voice. It came entirely from the comparison itself, from being scored against something in an out-group that's expected to win. We're wired to feel insecure in a situation of comparison. It's the same instinct that kicks in scrolling through someone else's highlight reel online: watching a life that looks better than yours makes you feel smaller, whether or not that was ever the point.
That's the real finding hiding inside this study. Robots occupy a strange spot in a workplace's social map. We instinctively sort coworkers into groups, us and them, the team and the new hire, the department and the contractors. Traditional industrial robots had their niche as tools, kept at arm's length, easy to categorize. The new generation, smarter and more capable, is a different story. They get sorted too, except they arrive pre-loaded with an assumption the rest of us don't carry: that they're supposed to outperform us. So a low score next to a robot doesn't read as a bad day. It reads as evidence for a story employees are already primed to believe.
There's a second finding worth sitting with. Zhao and her co-authors also asked participants to imagine losing to a fellow human coworker instead, no robot involved at all. Losing to a person still made people more anxious about robots. Not the coworker who beat them. The robot down the hall who hadn't been part of the comparison at all. Once someone starts to feel replaceable, the fear doesn't attach to whoever actually beat them. It attaches to whatever looks like the next, more permanent threat. Like a bush fire, it spreads on its own logic, and it's hard to contain once it starts.
So if a friendlier robot face doesn't fix any of this, what does. The researchers tested two options. Comparing employees to a fixed benchmark, rather than directly to a robot's score, softened the effect a little. But the strongest result came from something structural: when people were told they'd be evaluated and rewarded as part of a human-robot team, rather than judged head to head against the robot, their workplace attitudes held steady. Imagine that: same robot, same performance gap, just a different scoreboard. That was enough to stop the fire from spreading.
It's worth being honest about the limit here. Team-based framing kept commitment and satisfaction from dropping, but it didn't erase the underlying nervousness about being replaced one day. That worry didn't go anywhere. It just stopped showing up on the scorecard.
Which points to what leaders are actually managing when a robot joins a team. Not a design problem, how friendly does the robot look. A framing problem, how is performance being measured, and against whom. The moment a review or an incentive structure pits a person against a robot directly, you've built the comparison that produces the fear, no matter how the robot is dressed up. Structure the evaluation around the team instead, and the same robot, doing the same job, stops reading as a threat.
The robot was never the variable that mattered. The scoreboard was.
Source: Zhao, X., Zhang, X., & Leonardelli, G. J. (2025). When Your Co-Worker Is a Robot: Intergroup Performance Status and Its Consequences for Workplace Attitudes. European Journal of Social Psychology, 55(7), 1397-1419. Read the study →