Making AI Socially Responsible
An AI answer can sound neutral while carrying the assumptions of a biased record, a narrow source archive, or an institution that cannot be challenged. Who gets heard, and who bears the cost when the answer is wrong?
Artificial Social Intelligence: Making AI Socially Responsible examines how AI bias affects women and marginalized communities in employment, language technology, search, and everyday answers. Drawing on sociological theory and documented studies, the book follows bias from data collection through generation, institutional use, and repair. It also examines black boxes and the promise and limits of model cards, audits, standards, and regulation.
Five questions guide the reader: What can be known? Who is heard? What does the answer do? Who can decide? What happens after harm? Together, they offer a way to judge both the answer on the screen and the organizations behind it.
Written for students, educators, researchers, AI developers, and community advocates, this book makes the case for AI systems that can be questioned, corrected, and held responsible.
No comments:
Post a Comment