This research explores how Large Language Models are transforming security systems' ability to identify threats and unusual patterns.

• LLMs offer powerful capabilities for detecting anomalies beyond traditional methods
• Their advanced comprehension and generative abilities enable more sophisticated threat detection
• Integration of LLMs represents a paradigm shift in anomaly detection approaches
• Enhances reliability and trustworthiness of critical security systems

This research is critical for security professionals as it establishes how LLMs can strengthen defenses against novel threats and unknown attack patterns, potentially reducing false positives while improving detection accuracy.

Large Language Models for Anomaly and Out-of-Distribution Detection: A Survey