Bias refers to systematic errors or prejudices that can skew the outcomes of an AI system, often reflecting societal inequalities. Fairness aims to ensure AI treats all individuals and groups equitably, reducing discrimination and promoting justice.

This principle focuses on safeguarding personal data and ensuring that users’ information is not exploited or accessed without consent. Privacy protection is essential for maintaining trust and respecting individuals’ rights to confidentiality.

Explainability is the ability to understand and interpret how an AI system makes its decisions. This helps users and stakeholders comprehend the reasoning behind AI outcomes, fostering transparency and accountability.

Accessibility ensures that AI systems are usable and beneficial for all people, regardless of abilities, backgrounds, or technical knowledge. It promotes inclusivity by designing technology that everyone can engage with and benefit from.

Accountability means that those developing or deploying AI are responsible for its impacts, while transparency requires openness about how AI systems work. Together, these principles support ethical use by making AI operations visible and holding creators answerable.

Reliability refers to the consistency and dependability of an AI system’s performance. A reliable AI functions as expected, providing accurate and stable results even under different conditions or over time.

This principle ensures that AI systems are protected against threats and malicious attacks, minimising risks to users. Safety focuses on preventing harm from AI operations, while security safeguards data and system integrity.