Back to AI News
arxiv

SoK: Assessing the Evolution of LLM Capability in Software Vulnerability Research

This systematization of knowledge introduces a framework that rates LLM vulnerability-research capabilities on an autonomy scale modeled on driving-automation levels and qualifies claims by correctness, availability and evidence strength. It reports that agents discovered, validated and patched vulnerabilities in real-world software without human intervention before the restricted frontier models of 2026, with open-weight models later reaching comparable autonomy for discovery and validation. Across a decade of public vulnerability records, the authors observe that disclosed vulnerabilities more than doubled in 2026, with a growing share rated critical or high, while the time between report-to-fix and disclosure-to-attack stages has not shifted noticeably. The paper also grades proposed mitigations with an explicit rubric and derives guidance for stakeholder groups.

AI Brief

It provides a structured, evidence-graded view of how autonomous LLM security agents are changing vulnerability discovery and disclosure, useful for security and AI policy readers.

arXiv:2611.26601v1 Announce Type: cross Abstract: LLM-based agents are becoming increasingly capable at discovering, validating and patching vulnerabilities in real software, and vendors and researchers report that this is already changing how vulnerabilities are found and fixed. These claims, however, are scattered across papers, vendor reports and disclosures and are rarely comparable. We introduce a framework that rates these capabilities on an autonomy scale modeled on the levels of driving automation, and qualifies each claim by its correctness, its availability and the strength of its evidence. We use this framework to systematize the published evidence on LLM capability in vulnerability research. We find that agents discovered, validated, and patched vulnerabilities in real-world software without human intervention before the restricted frontier models of 2026, and that open-weight models subsequently attained comparable autonomy for discovery and validation. We then measure the consequences of evolving LLM security capabilities in a decade of public vulnerability records, advisories and exploitation data. In 2026 the number of disclosed vulnerabilities more than doubled, with an increasing share rated critical or high, and severe bugs that had survived a decade or more in widely used code were reported several times more often than before. Yet the time between the stages of a vulnerability's life, from report to fix and from disclosure to attack, has not shifted noticeably so far. We grade the mitigations proposed in the literature with an explicit rubric and draw guidance for each group of stakeholders from the best-supported ones.
Be the first to rate this article.

Comments & Questions (0)

Captcha
Please be respectful — let's keep the conversation friendly.

No comments yet

Be the first to comment!