AI security is emerging and it necessary for organization and AI practitioners to incorporate secure way of developing and adopting AI systems. This means at each stage of SDLC there needs to controls and best practices to be followed specific to AI apart from traditional application security practices (Secure SDLC).
SDLC Stage vs AI Security Controls:
Design - Risk Assessments / Threat Modelling AI systems Development - Asset Inventory / tracking , protect sensitive data and supply chain security
Deployment - Secure MLOps, Incident Management, AI Security assessments, Secure Infrastructure
Operate and Maintain - Secure Logging and Monitoring, Patch management.
References and Resources will be shared.
By the end of this talk, the audience will have a clear understanding of how to approach AI security and will be equipped to develop guidelines that ensure the security of AI systems throughout their lifecycle.
