The UK government has published a Code of Practice for the Cyber Security of AI, providing voluntary cyber security requirements for the lifecycle of AI. The Code defines stakeholders in the AI supply chain, including developers, system operators, and end-users. The Code is broken down into 13 principles, each containing provisions for compliance. These principles aim to raise awareness of AI security threats and risks.
The Code’s principles cover key areas such as designing AI systems for security, evaluating threats and managing risks, and securing infrastructure and supply chains. The accompanying Implementation Guide provides guidance on meeting the cyber security requirements outlined in the Code. The Code and Guide will also be used as the basis for a new global standard.
