A NIST CSF-based cybersecurity profile for AI and machine learning systems in transportation networks

The increasing adoption of artificial intelligence (AI) and machine learning (ML) systems in transportation networks, including intelligent transportation systems, autonomous vehicles, and logistics platforms, has introduced new cybersecurity risks that are not fully addressed by existing frameworks. Adversarial attacks such as evasion, poisoning, and privacy attacks can directly impact transportation safety, disrupt operations, and compromise critical infrastructure. While the NIST Cybersecurity Framework (CSF) 2.0 provides a widely used struc-ture for managing cybersecurity risks, it does not define AI- or ML-specific categories. Similarly, the NIST AI Risk Management Framework focuses mainly on governance and trustworthiness, with limited oper-ational guidance for handling adversarial threats in deployed systems. This paper proposes a CSF-based cybersecurity profile tailored for AI and ML systems in transportation environments. The proposed profile extends the CSF functions—Govern, Identify, Protect, Detect, Respond, and Recover—with AI-specific categories and subcategories that address adversarial ML risks. The framework maps AI threats to operational cybersecurity processes, enabling structured handling of incidents across transportation systems and networks. The evaluation demonstrates that the proposed profile improves threat coverage, reduces exposure time, and supports more consistent handling of adversarial events. The proposed approach provides a practical and structured solution for enhancing the security, reliability, and resilience of AI-driven transportation systems.