When LLMs Meet Cybersecurity: A Systematic Literature Review
A systematic literature review indexing more than 861 papers on large language models applied to cybersecurity, organized around domain-specific LLMs, defensive and offensive applications, and future research directions.
When LLMs Meet Cybersecurity is a systematic literature review indexing more than 861 papers on large language models applied to cybersecurity, organized around three research questions: building domain-specific LLMs, applying LLMs within cybersecurity, and future directions for the field. It reports each paper's publication venue and tracks how many new papers are added with every update.
Papers are broken down by category, including evaluation benchmarks and fine-tuned domain LLMs for building domain-specific models; LLM-assisted defense, vulnerability detection, LLM-assisted attack, program and vulnerability repair, threat intelligence, fuzzing and insecure code generation for applying LLMs to security work; and an emerging category on LLM agents for cybersecurity for future directions. The full literature review, with paper counts and short summaries for each category, is kept in a separate linked file for readability.
It is regularly updated with dated entries noting how many papers were added at each revision.