CyberzSOC

Supply chain risks and mitigations 2 Artificial intelligence and machine learning Supply chain risks and mitigations Artificial intelligence and machine learning Supply chain risks and mitigations 3 Artificial intelligence (AI) and machine with this, risks outlined in this guidance are learning (ML) systems allow organisations to mapped to the National Institute of Standards improve their efficiency in many areas. These and Technology’s (NIST) Adversarial Machine systems can help inform decisions, streamline Learning (AML) taxonomy at csrc.nist.gov, processes and improve customer experience. where applicable. This guidance in general maps to MITRE’s Adversarial Threat Landscape Adopting AI and ML systems introduces unique for Artificial-Intelligence Systems (ATLAS) supply chain risks, which can threaten the framework under AI Supply Chain Compromise at cyber security of an organisation if not securely atlas.mitre.org. Using pre-trained models and thirdInformation section at the end of this guidance. party datasets is beneficial, but it can also bring existing compromise and supply chain risks. This guidance focuses on aspects unique to AI Organisations should know what to look out and ML, though general cyber supply chain risk for when developing or incorporating AI and management still applies. Wider AI development at cyber.gov.