Degree of Transparency and Explainability
Transparency is the characteristic of a system that describes the degree to which appropriate information about the system is communicated to relevant stakeholders, whereas explainability describes the property of an AI system to express important factors influencing the results of the AI system in a way that is understandable for humans....Information about the model underlying the decision-making process is relevant for transparency. Systems with a low degree of transparency can pose risks in terms of their fairness, security and accountability.
ENTITY
2 - AI
INTENT
2 - Unintentional
TIMING
2 - Post-deployment
Risk ID
mit183
Domain lineage
7. AI System Safety, Failures, & Limitations
7.4 > Lack of transparency or interpretability
Mitigation strategy
1. Establish a comprehensive AI Governance Framework and Policy: Implement a formal framework, such as one aligned with the NIST AI Risk Management Framework or ISO/IEC 42001, to mandate "Transparency by Design" throughout the AI lifecycle. This strategic governance action defines the organizational risk appetite, assigns clear accountability for transparency and compliance, and provides the foundation for all subsequent technical and procedural mitigations. 2. Integrate Explainable AI (XAI) Techniques: Adopt model-agnostic (e.g., LIME, SHAP) and model-specific methods to provide both global (model-level) and local (instance-level) explanations for all AI decisions. The explanation must be tailored to the stakeholder's needs—simple, plain-language justifications for end-users and detailed algorithmic insights for developers—to ensure human comprehensibility without compromising model performance. 3. Institute Continuous Auditing, Validation, and Monitoring: Mandate routine, repeatable risk assessments, stress testing (including adversarial red teaming), and automated data validation pipelines to actively monitor the AI system post-deployment. This continuous process ensures that the initial levels of transparency and interpretability are maintained, proactively detecting and mitigating any emerging systemic biases, data drift, or vulnerabilities that could undermine fairness and accountability.