A new study finds that AI anxiety can be triggered just by reading the right story, not by an actual layoff.
Researchers built the work on fear acquisition theory and the technology acceptance model, then ran two studies to test what actually drives AI job replacement anxiety. In the first, a randomized experiment with 316 participants, exposing people to narratives that emphasized AI acting with agency and taking control away from humans measurably increased their anxiety about losing a job to AI. That effect held even after the researchers controlled for participants' general self-evaluations, and it was stronger among people who already saw AI as genuinely useful, though not among those who simply found it easy to use. A second study, an observational survey of 995 people, replicated the core finding and backed up the moderating role of perceived usefulness.
The real contribution here is the causal piece: this is the first evidence that AI job anxiety can be triggered vicariously, through a narrative, rather than only through direct experience of displacement. That reframes a debate usually fought over labor statistics and automation forecasts into one about language and framing: how AI capabilities get described in the press, in marketing, and in casual conversation may do as much to shape fear as any actual job-loss numbers.
Worth remembering the next time a vendor's demo video shows an AI system confidently taking over a task end to end: the framing itself, independent of whether the tool works as advertised, is doing psychological work on the audience.