AI Assurance
The process of measuring, evaluating, and communicating the trustworthiness of AI systems.
What it triggers
- In the United Kingdom, the Introduction to AI assurance sets out voluntary risk frameworks and guides organizations and practitioners on evaluating and communicating the trustworthiness of AI systems.source
- In the United Kingdom, the Introduction to AI assurance describes AI assurance as a systematic process of gathering reliable, standardized, and accessible evidence about an AI system's capabilities, intended functionality, limitations, and risks to build justified trust.source
Definitions (2)
AI assurance refers to the systematic process of gathering reliable, standardized, and accessible evidence about an AI system's capabilities, intended functionality, limitations, and risks. It aims to build justified trust by ensuring trustworthiness is established and effectively communicated to stakeholders.
AI assurance is the practical assessment and management of an AI use case through impact assessment, consultation, testing and monitoring. It describes the guide’s implementation approach rather than a statutory definition.
Questions and answers
What is the purpose of AI assurance according to the UK guideline?
AI assurance aims to build justified trust by ensuring trustworthiness is established and effectively communicated to stakeholders. It involves gathering reliable, standardized, and accessible evidence about an AI system's capabilities, intended functionality, limitations, and risks.
Is the UK Introduction to AI assurance a binding requirement?
No, it is a guideline that is not binding. It guides UK organizations and practitioners on evaluating and communicating the trustworthiness of AI systems and sets out voluntary risk frameworks.
Who published the Introduction to AI assurance?
It was published by the Department for Science, Innovation and Technology.