Core Concepts
Bias
AI bias occurs when a model systematically produces unfair or skewed outputs, often because the training data reflects historical inequalities or under-represents certain groups. It manifests in hiring tools, facial recognition, loan approvals, and many other applications.
Addressing bias requires diverse training data, fairness-aware model evaluation, and ongoing auditing in deployment. Regulatory bodies increasingly require bias assessments for AI systems used in consequential decisions.
Authority Links
Related Terms
Core Concepts
Explainable AI (XAI)
AI systems that provide transparent insights into their decision-making processes.
Techniques & Methods
AI Alignment
The research field and engineering practice of building AI systems that reliably pursue goals humans actually want, remain controllable, and avoid harmful side effects — operationalized through RLHF, Constitutional AI, evaluations, and interpretability.
Miscellaneous
Training Data
The labeled or unlabeled dataset used to fit a model's parameters during the learning process.
Techniques & Methods
Evaluation Metrics
Quantitative measures used to assess how well an AI model performs on a task.