Artificial Intelligence promises inclusion but often delivers invisibility.
The Bias Pipeline examines how systems of data, design, and policy quietly decide who becomes legible to machines and who remains excluded.
This site continues the conversation that began with my talk “Prototype — Accessible to Whom? Legible to What?” It gathers essays, notes, and resources on AI, accessibility, and disability-led design.
The aim is threefold:
To expose how bias travels from data into decisions.
To advocate for disability-led design and accessible technology.
To push policy actors — including NITI Aayog — to align India’s AI strategy with the UNCRPD, the RPwD Act, and the Rajiv Raturi judgment.
Accessibility here is not an afterthought; it is an integral part of the architecture.
When disability leads, AI learns to listen.