Category · 10 terms
AI Governance, Risk & Safety
defined clearly.
Bias, fairness, explainability, regulation, red-teaming — the trust stack.
AI Ethics
The philosophical and applied study of moral questions raised by AI — what we ought to build, deploy, and forbid, and on what principles.
AI Governance
The frameworks, policies, and controls that organizations apply to ensure AI systems are deployed safely, ethically, legally, and aligned with business goals.
AI Incident
An event in which an AI system causes or nearly causes harm — recorded and analyzed so the field learns from failures the way aviation does.
AI Red Teaming
Structured adversarial testing of an AI system to discover failure modes, jailbreaks, prompt injections, and harmful outputs before deployment.
Algorithmic Fairness
A research area concerned with whether ML systems produce equitable outcomes across protected groups, and the mathematical and policy choices involved.
EU AI Act
The European Union's comprehensive AI regulation (in force 2024-2026), the first major legal framework to classify AI systems by risk and impose obligations accordingly.
Jailbreak (LLM)
A prompt or technique that bypasses an LLM's safety training, eliciting outputs the model is supposed to refuse.
Model Card
A standardized documentation artifact that describes a trained model — its purpose, training data, evaluation metrics, intended use, and known limitations.
Prompt Injection
An attack where adversarial input — often hidden in retrieved or tool-returned content — overrides the developer's instructions to an LLM.
Responsible AI
A practice of designing, building, and deploying AI systems that are fair, transparent, accountable, safe, and respectful of user privacy and rights.