algorithmic bias
Automated unfair outcomes
Algorithmic bias refers to systematic errors in computer systems that produce unfair outcomes, typically favoring or disadvantaging particular groups.
The bias often reflects prejudices in training data or the assumptions of developers: The hiring algorithm showed algorithmic bias against women.
The term emerged as machine learning systems began making consequential decisions about loans, criminal sentencing, and employment.
- predictive policingAlgorithm-guided law enforcement deployment
- discriminationUnjust differential treatment
- digital divideUnequal access to computers/internet
- disparate impactDiscrimination through neutral practices
- protected classCategory with anti-discrimination shields
transparency · AI ethics · redlining · machine learning · proxy · recidivism · risk assessment · training data · feedback loops · filter bubble · inequality · data privacy · legal challenges · automation · human oversight · privacy concerns · unintended consequences · hiring
- AI ethicsrobot rights, AI rights, moral agency, human oversight, explainability
- biaseshalo effect, horn effect
- Digital Ethicsprivacy, data privacy, cyberbullying, misinformation, disinformation
- financial biasredlining, credit denial, loan denial, fair lending, interest rates
- tech ethicsAI ethics, data privacy, deepfakes, digital rights, tech addiction
- tech policydata privacy, net neutrality, tech policy, digital rights, content moderation