1 - The obligation of an organisation or individual to take responsibility for the outcomes of AI systems.

2 - The characteristic of an AI system that supports holding the operator and/or provider to account.