The robot didn't take a single job. Nobody was made redundant, nobody's hours were cut, and management said so, repeatedly. Six weeks later, engagement scores dropped anyway.
Here's what we normally assume. People resist robots because they're afraid of losing their jobs. So we lead with the reassurance: your role is safe, the robot handles the repetitive work, you'll be freed up for better things. Say it early, say it often, and the resistance should fade.
Turns out that answers a question people were only half asking.
Xian Zhao, with colleagues Zhang and Leonardelli, ran five experiments published in the European Journal of Social Psychology, using both scenarios and simulated work environments. They wanted to know what happens inside people when a robot performs alongside them. The finding: when people perceive a robot as outperforming humans, their commitment to the organisation and their satisfaction with the job both drop. Not because the robot threatens their employment directly. Because a comparison has been opened, human team versus machine, and their side is losing. You know this from social media. You look at people living beautiful lives and feel worse for the comparison. Same effect here.
Psychologists call this intergroup comparison, and you already know it from ordinary work life. It's the sales team checking the other region's numbers. It's the night shift muttering about the day shift. We read our worth partly through how our group stacks up against the group next door. The robot walks in and becomes the group next door. Except this rival doesn't tire, doesn't take leave, and improves with every software update.
Two details in the study are worth sitting with.
First, the researchers tested whether it mattered how humanlike the robot was. Surprisingly, it didn't. The friendly one with eyes and the industrial arm produced the same effect. So the fix is not a cuter robot.
Second, and this is the clever part, they tested people against a plain performance benchmark, a number on a wall, set at the same level as the robot's output. No threat. No drop in attitudes. Falling short of a standard doesn't wound us. Being outperformed by a rival group does. It's not the gap that feels painful. It's the loser's position in a contest nobody announced.
The study's explanation is plain: negative performance comparisons can imply your group is replaceable. That's the sentence under the sentence when a worker looks at the robot's numbers. Not "I might be fired." More like "we might not matter."
Which explains something many managers find baffling. You've guaranteed the jobs. You've said the robot does the boring work. And morale still sags, quietly, in ways nobody can articulate in the town hall. The wound was never employment. It was status. And status wounds don't show up in the questions people ask. They show up in commitment, which is exactly where the study found the damage.
Now look at how most workplaces run a rollout. The robot's throughput goes up on a dashboard. Someone circulates a chart, human pick rates and machine pick rates, side by side. It's meant as transparency. Read through the intergroup lens, it's a league table where one team was born to lose, published weekly.
The researchers point at a repair that costs nothing: team-based measurement. Review the performance of the unit, humans and robot together, rather than humans versus robot. Same data, different frame. The robot stops being the rival group and becomes something closer to equipment the team wins with. Whether the machine is on our side of the comparison, it turns out, matters more than what the machine can do.
So here's the lens shift. The next time a rollout meets resistance that reassurance can't touch, stop asking whether people feel their jobs are safe. Ask whether you've accidentally put them in a contest with the machine. People can share a floor with a robot. What they can't do is lose to one, every week, in public.
Source: Zhao, X., Zhang, & 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 (Wiley). Read the study →