Open data for "Willing to Be Read? A Systematic Review of Public Acceptance of Emotion Recognition Technologies and the Contexts Prohibited by the EU AI Act

This deposit contains the search log, the eligibility log, the coded extraction dataset and a dependency-free Python script for the vote-counting analysis (exact sign and Fisher tests) of a PRISMA 2020 systematic review. The review covers 27 peer-reviewed reports (25 samples, 2016–2026) on public acceptance of emotion recognition technologies (emotional AI) across deployment contexts and modalities, and compares the results with the EU AI Act's prohibition of emotion recognition in workplaces and education. Main findings: acceptance was predominantly negative in 11 of 22 codable reports and positive in 4. All workplace and hiring assessments were negative. Respondents asked to imagine adopting a tool for themselves were never predominantly negative, whereas respondents framed as data subjects were negative in 11 of 17 reports. The review proposes the Beneficiary–Asymmetry–Control (BAC) framework.