(TR05) (INCYBER) How do we assess and “standardize” AI safety?

AI safety assessment and standardization are crucial issues in a world that is becoming increasingly dependent on AI. How do we effectively assess the risks associated with using AI, such as biases or technical vulnerabilities? What standardized criteria and processes should we put in place to guarantee the safe development and deployment of AI? How should existing standards (e.g. ISO 27001 and ISO 9001) be adapted to keep pace with the growth of AI?