The UK’s National Cyber Security Centre has joined with other international government agencies, with responsibility for cyber security, from around the world to create guidelines for the secure developement of AI systems. These guideline cover, data poisoning, input manipulation, privacy and IP threats, re-identification of anonomised data, model stealing and data exfiltration, team and contractor considerations among other top AI threats and risks.
Guidelines for secure AI system development (ncsc.gov.uk)
It is worth reading even if you do not work in AI – as AI is going to affest us all.
Clive Catton MSc (Cyber Security) – by-line and other articles