On managing vulnerabilities in AI/ML systems

TL;DR

This paper examines how integrating CVE identifiers for ML vulnerabilities could improve communication between research and operational security communities, addressing current gaps in vulnerability management for AI/ML systems.

Contribution

It proposes a thought experiment on assigning CVE-IDs to ML flaws, highlighting potential improvements in vulnerability management and cross-community communication.

Findings

CVE-IDs could bridge research and operational security communities.

Current gap in communication between ML research and deployment.

Potential for CVE-IDs to improve vulnerability reporting and response.

Abstract

This paper explores how the current paradigm of vulnerability management might adapt to include machine learning systems through a thought experiment: what if flaws in machine learning (ML) were assigned Common Vulnerabilities and Exposures (CVE) identifiers (CVE-IDs)? We consider both ML algorithms and model objects. The hypothetical scenario is structured around exploring the changes to the six areas of vulnerability management: discovery, report intake, analysis, coordination, disclosure, and response. While algorithm flaws are well-known in the academic research community, there is no apparent clear line of communication between this research community and the operational communities that deploy and manage systems that use ML. The thought experiments identify some ways in which CVE-IDs may establish some useful lines of communication between these two communities. In particular, it would start to introduce the research community to operational security concepts, which appears to be a gap left by existing efforts.