Leveraging large language models for multi-domain deception detection
Abstract
Deception detection falls in the category of study that is applicable in various fields, which include security, forensics, journalism, and social media monitoring. These traditional methods rely mainly on human judgment or rule-based systems and are generally not flexible or scalable. The work focuses on large language model capabilities for detecting misleading language through advanced contextual reasoning and natural language understanding abilities. We examine several linguistic and psychological cues that suggest deception, such as vagueness, excessive detail, inconsistencies in emotions, and disfluencies.
Fields of Study & Subjects
Misinformation and Its Impacts
Deception detection and forensic psychology
Authorship Attribution and Profiling
Publication details
- Journal / source
- AIP conference proceedings
- Volume
- 3393
- Pages
- 020029–020029
- Publication date
- 1 January 2026
- Language
- EN