26/08/2026
🔍 Most research on AI job anxiety asks who is at risk: which occupations are exposed, which tasks are automatable. Far less attention goes to how AI is talked about. We tested that directly, and our is now available on arXiv: https://arxiv.org/abs/2608.14609
🧪 In a preregistered (N = 316), participants watched compilations of short social media videos. One condition emphasised operating and deciding independently of human oversight; the other emphasised human control, regulation, and governance. None of the videos mentioned job loss. Turns out that sole exposure to the AI-control condition increased job replacement anxiety.
📊 Then, a survey (N = 995) reproduced the association outside the lab, adding to external validity of the results.
⚠️ Two findings ran against our predictions. 1) Perceived usefulness of AI did not buffer anxiety. People who considered AI more useful were more anxious about replacement, regardless of which narrative they saw. 2) Perceived ease of use did nothing at all. Effort-related appraisals may simply be irrelevant when someone is not a user of a technology but a potential subject of its consequences.
💡 The practical implication is uncomfortable. Job replacement anxiety can be triggered by how AI agency is framed, before any measurable labour-market change occurs. Communication about AI is not neutral with respect to employee well-being.
Data, materials, preregistration: https://osf.io/yrjg4/
As artificial intelligence (AI) rapidly diffuses and concerns about job displacement intensify, the psychological mechanisms underlying AI job replacement anxiety remain insufficiently understood. Drawing on Integrated Fear Acquisition Theory and the Technology Acceptance Model, the present research...