Eticas.ai has been evaluating AI systems in production since 2012. This is where we share what we’ve learned. You’ll find case studies from client work, practical guides on AI governance and evaluation, reports on emerging risks, interviews with practitioners, and a glossary of the concepts that matter most in the field. Whether you build AI, deploy it, or are accountable for its outcomes, this is the evidence base we work from.
Algorithmic Bias
Algorithmic bias occurs when an AI system produces systematically different outcomes for different groups of people based on protected characteristics like race, gender, age, or socioeconomic status. It can originate in training data, in feature engineering that introduces proxy variables, or in threshold decisions that were never tested across the full population the system serves.