Detection of malicious software using large language modelsTivadar, Martina (Avtor) Karakatič, Sašo (Mentor) malwarelarge language modelsdetectionThis thesis examines the success rate of large language models (LLM) in detecting macOS malware through Endpoint Security logs. A literature review and 144 experiments with three ChatGPT variants and six prompt types evaluated accuracy, precision, recall, specificity, and F1-score. Results show that prompt wording is crucial: zero-shot and chain-of-thought prompts performed best, while conservative prompts minimized false positives but missed threats. GPT-4o and o1 outperformed o4-mini but showed similar results. Findings suggest LLMs can support, but not replace, traditional detection, with prompt design proving as important as model choice.[M. Tivadar]20252025-09-16 17:42:31Magistrsko delo/naloga95400sl