Cyber Security Requirements for AI Models and Systems principles to Generative AI; Preventing Harm from Deepfakes (ETSI TS 104 275)
The work item will be a domain-specific focus application of the EN 304 223 principle for base cybersecurity requirements for AI models and systems to Generative AI. The TS will focus on the subject of deepfakes, which includes the issues of misinformation, disinformation, loss of confidentiality, copyright, IPR and the training/source material for LLMs. Also, this TS will include more prescriptive specifications for this domain extension to EN 304 223 when necessary.