Online job scams: Unveiling the impact of overconfidence, digital literacy, and algorithmic literacy on user susceptibility to false job advertisements

This article examines false online job ads and user susceptibility by drawing on three research areas: automated deception detection, cognitive bias (Dunning–Kruger effect), and digital and algorithmic literacy. Leveraging a data set of 17,879 ads, we develop machine learning models to distinguish false from legitimate ads and survey a representative US sample ( N = 635) to assess user susceptibility. Results show that ad veracity is predicted by visual (e.g. company logo) and linguistic cues (e.g. “team,” ‘we’) that signal credibility. Survey findings indicate a strong Dunning–Kruger effect: overconfident individuals were less accurate in detecting false ads, perceived them as more credible, and were more likely to share them. Literacy effects were complex: while general Internet skills improved detection accuracy, algorithmic literacy had more nuanced influences on ad perceptions. These results highlight the psychological mechanisms that contribute to the spread of false information in digital communication, deepening our understanding of online deception.