AI Assisted Forensic Linguistics: Methods, Potentials and Challenges in the Analysis of Malicious Texts

Recently, Forensic Linguistics has faced the challenge of consolidating its status as a rigorous scientific discipline. One of its current priorities is the development and validation of reliable methods and tools that can support expert analyses in legal contexts. In this regard, the emergence of language models based on Artificial Intelligence opens new horizons for research and application. The aim of this article is to explore the potential of ChatGPT as assistant in the forensic analysis of malicious texts. Building on a corpus previously examined, the study compares the findings derived from human discourse analysis, supported by tools such as Tropes and Seoscout, with those obtained through interaction with ChatGPT. The analysis evaluates the model’s capacity to detect key linguistic indicators, including threat markers, persuasive strategies, argumentative structures, and stylistic patterns. The usability as a complementary resource in forensic linguistic investigations is discussed, along with its limitations. The article includes a critical reflection on the implications of using AI technologies in Computational Forensic Linguistics, highlighting both their promise and the need for caution in their integration into expert practice.