AI.info
Responsible AI
Governance as sociotechnical work — stakeholders, foreseeable misuse, and a system followed from proposal to retirement.
- Responsible AI as Sociotechnical Governance
- Stakeholders, Affected People, and Power
- Intended Purpose, Context of Use, and Foreseeable Misuse
- Harm Taxonomy and Risk Scenarios
- Risk Appetite, Acceptance Criteria, and Stop Rules
- AI System Inventory, Scope, and Lifecycle Classification
- Accountability, Decision Rights, and Independent Challenge
- AI Literacy, Competence, and Role-Based Training
- Policies, Standards, Controls, and Evidence
- Human Rights, Democratic Values, and the Rule of Law
- Algorithmic Impact Assessment and Proportionality
- Accessibility and Inclusive AI Design
- Children, Vulnerable Groups, and High-Dependency Contexts
- Labor, Environmental, and Supply-Chain Impacts
- Dual Use, Misuse, and Societal Externalities
- Bias Is More Than Biased Data
- Sampling, Representation, and Measurement Bias
- Labels, Proxies, and Historical Decision Bias
- Group Performance and Intersectional Evaluation
- Fairness Criteria and Incompatible Goals
- Fairness Mitigation Across Data, Models, Decisions, and Institutions
- Causal and Counterfactual Fairness
- Thresholds, Resource Allocation, and Fairness in Operations
- Fairness Monitoring, Complaints, and Remediation
- Privacy Threat Modeling for AI Systems
- Data Minimization, Purpose Limitation, Consent, and Retention
- Automated Decisions, Profiling, and Individual Rights
- Memorization, Membership Inference, Model Inversion, and Extraction
- Differential Privacy and Privacy Budgets
- Federated Learning and Secure Aggregation
- MPC, Homomorphic Encryption, Trusted Execution, and PET Selection
- Synthetic Data, Privacy-Preserving Evaluation, and Residual Risk
- Transparency by Audience and Decision
- Explainability: Intrinsic, Local, Global, and Counterfactual
- Faithfulness, Stability, and the Limits of Explanations
- Datasheets, Model Cards, and System Cards
- Disclosure, AI-Generated Content, and Provenance
- Human Oversight, Automation Bias, and Workload
- Contestability, Appeal, Recourse, and Redress
- Safety Cases, Hazard Analysis, and Misuse Cases
- Robustness, Distribution Shift, and Operational Boundaries
- AI Security and Adversarial Machine Learning Governance
- Red Teaming, Evaluation Governance, and Independent Challenge
- AI Incident Reporting, Corrective Action, and Learning
- Third-Party, Vendor, Open-Source, and Supply-Chain Governance
- Global AI Governance: Principles, Treaties, and Interoperability
- EU AI Act: Roles, Scope, and Risk Categories
- EU AI Act: Obligations, GPAI, Transparency, and the 2026 Timeline
- GDPR, DPIAs, and Automated Decision-Making
- AI Management Systems: NIST AI RMF, ISO/IEC 42001, and Impact Assessment
- Sectoral Governance, Board Oversight, and Speak-Up Culture
- Responsible AI Capstone: Govern a High-Impact System From Proposal to Retirement