Trustworthy AI Characteristics

Seven interrelated characteristics that together constitute AI trustworthiness:

  1. Valid and Reliable: System performs as required under expected conditions; accuracy and robustness are demonstrated
  2. Safe: System does not endanger human life, health, property, or environment under defined conditions
  3. Secure and Resilient: System maintains confidentiality, integrity, and availability; withstands adverse events
  4. Accountable and Transparent: Information about system and outputs is available to those interacting with it
  5. Explainable and Interpretable: Mechanisms underlying operation can be represented; outputs have meaning in context
  6. Privacy-Enhanced: Practices safeguard human autonomy, identity, and dignity
  7. Fair with Harmful Bias Managed: Systemic, computational, and human-cognitive biases are identified and mitigated

Valid & Reliable is positioned as foundational, other characteristics build upon it. Accountable & Transparent relates to all other characteristics.

Creating trustworthy AI requires balancing each characteristic based on context of use. Tradeoffs are unavoidable.

Related: 05-atom—validity-reliability-foundation