Feminist AI is artificial intelligence built on the premise that technical systems inherit the social structures of the people who design them. Rather than treating fairness as a downstream correction, feminist AI treats equity as a design constraint from the first line of code. It asks not only whether a system is accurate, but whether it is just, and for whom.
Automated systems now allocate credit, screen job applicants, triage healthcare, and shape what information reaches whom. When these systems are trained on historical data, they tend to reproduce historical exclusion at scale and at speed. A feminist lens exposes that inheritance and interrupts it before deployment rather than after harm.
The A+ Alliance for Inclusive Algorithms treats feminist AI as pro-social by design, oriented toward correcting inequity rather than merely optimizing efficiency. This orientation runs through the Alliance’s origin: the 2019 Affirmative Action for Algorithms declaration in Zürich, and through its ongoing work with the Feminist AI Research network and the Gender & AI Innovation Collective.
