When Machines Mislead: Trust, Cognitive Processing, and Belief in AI-Labeled Misinformation

Generative artificial intelligence (AI) has increased the speed and scale at which plausible text and other media can be produced, raising questions about how people evaluate information explicitly attributed to AI. This integrative narrative review brings together research on misinformation judgments, online credibility cues, trust in automation, dual-process accounts of reasoning, and Need for Cognition (NFC). The literature suggests that people may use source labels as one cue among several when judging credibility, while trust in AI can shape willingness to rely on AI-generated information. Analytic reasoning is associated with better discrimination between true and false news, but NFC is a motivational disposition toward effortful thinking and should not be treated as a direct measure of analytic ability or heuristic use. Evidence also cautions against assuming that AI disclosure uniformly increases or decreases credibility: outcomes may depend on the message, the source cue, prior expectations, and the task. The review proposes a cautious framework in which AI attribution is interpreted through individual beliefs about AI and evaluated alongside message-level evidence. Direct evidence on the joint roles of trust in AI and NFC in responses to AI-labeled misinformation remains limited. Future research should measure perceived accuracy and truth discernment separately, distinguish source trust from message accuracy, and test these relationships across culturally diverse samples, including India.