AI Accountability (ACC)

Accountability refers both to a characteristic of organizations and to a system property:
  1. 1.
    Organizational accountability means that an organization takes responsibility for its decisions and actions by explaining them and being answerable for them to the governing body, to legal authorities and more broadly to stakeholders.
  2. 2.
    System accountability relates to being able to trace the decisions and actions of an entity to that entity.
The use of AI can change existing accountability frameworks. Where previously persons performed actions for which they would be held accountable, such actions can now be fully or partially performed by AI systems. Who would be accountable in this case is an ongoing consideration by regulators. Developers and users of AI systems should be aware of the related legislation in the countries where the system is brought onto the market and used.
Controls related to this risk category are listed as below:
  • ACC 01 - Intended Purpose of AI Use
  • ACC 02 - Accountability Framework
  • ACC 03 - Logging Capabilities
  • ACC 04 - Logging Traceability
  • ACC 05 - Full Track of Mitigation Measures
  • ACC 06 - TEVV Documentation
  • ACC 07 - Tools for TEVV
  • ACC 08 - Evaluations of Human Subjects