Fair-Code
An AI fairness and algorithmic bias awareness initiative combining open-source development, ethical AI education, and public campaigns on real-world bias — COMPAS sentencing, automated hiring systems, and credit scoring.
- AI Fairness
- Open Source
- Ethics
- Research
- Education
01 / Reach
- 21K+
- Views
- 40
- Stars
- 16
- Forks
- 8
- Watching
- 15
- Countries
- 1,100+
- Interactions
Sustained public engagement and continued growth across 13 countries.
02 / The gap
Problem
Algorithmic bias harms millions of people, yet most of the public — and a surprising number of engineers — don't know it exists in the systems they rely on every day.
Approach
An open platform documenting real, documented bias cases, paired with tools for fairness analysis and public education campaigns that make the problem legible to non-specialists.
03 / What it covers
- COMPAS sentencing — how recidivism-risk scoring produced disparate outcomes in the criminal justice system.
- Automated hiring — where résumé screening and ranking models learn and amplify historical hiring patterns.
- Credit scoring — how proxy variables let models reproduce discrimination without ever reading a protected attribute.
- Explainers, in the open — written for a general audience and published as open source, so anyone can read, reuse, or contribute.
04 / Why it matters
Fairness work fails when it stays inside research papers. Fair-Code exists to move it into public view: documenting cases in plain language, keeping every explainer open source, and treating "who does this harm, and by how much?" as the first question rather than the last.