Standard

ETSI TS 104 223

Published

Singles purchase not accessible

This standard can not be purchased as a single sales.

Abstract

The present document defines baseline security requirements for AI models and systems. This includes systems that incorporate deep neural networks, such as generative AI. For consistency, the term "AI systems" throughout the present document when framing the scope of provisions and "AI security" which is considered a subset of cyber security. The present document is not designed for academics who are creating and testing AI systems only for research purposes (AI systems which are not going to be deployed).  The present document separates principles and requirements into five phases. These are secure design, secure development, secure deployment, secure maintenance and secure end of life. Relevant standards and publications are signposted at the start of each principle to highlight links between the various documents and the present document. This is not an exhaustive list.  NOTE: The principles can also be mapped to the life cycle stages described in ISO/IEC 22989 [i.3]. The secure design and development principles can be applied during the Design and development life cycle stage. Similarly, the secure deployment principles can be applied during the Deployment stage, secure maintenance to the Operations and monitoring stage and secure end of life during the Retirement stage.

Document information

  • Standard from ETSI_AC
  • Published:
  • Version: 0
  • Document type: IS
  • Additional information
    • V1.1.1