Abstract
Equitable AI refers to developing and using artificial intelligence systems in a manner that is responsible, fair, and inclusive.
1. Responsible: AI technologies are designed, developed, deployed,
and used in ways that are ethical, fair, and beneficial to all members
of society.
2. Fair: AI technologies do not reinforce existing biases that can result
in discrimination against certain groups.
3. Inclusive: AI technologies are easily accessible by all, including
those in remote or underserved areas.
Researchers of the Alan Turing Institute design and develop AI technologies for a variety of applications including finance, health, security, infrastructure, environment, and general-capability development. In
all these applications, one of the main objectives is delivering equitable outcomes for a diverse community. The Turing’s Gates Foundation– Our project, Trustworthy Digital Infrastructure for Identity Systems, focuses on "trustworthiness" which encompasses responsible, fair and inclusive development as well as robustness, resilience, safety, security, privacy, accountability and transparency.
funded project aims to inform and develop technologies for trustworthy and equitable digital infrastructure to enable access to critical public
services for all, especially in underserved areas.
An example digital public infrastructure (DPI) is national identity (ID) schemes. Identity is a complex and multi-component construct with strong cultural and societal values [1]. It is also the gateway to
accessing vital public services such as enrollment in the education system, healthcare, and social benefits. The basic role of an ID system
is answering three questions [2]:
(1) Who are you? (Identification),
(2)Are you who you claim to be? (Authentication),
(3) Are you authorized/eligible? (Authorization).
Considering these functions of the ID systems, it is essential to provide fair and equal access for all citizens. Our commitment to equitable AI is rooted in a belief that technology should serve as a tool for social good, bridging gaps rather than widening them. The importance of equitable AI extends beyond ethical imperatives; it is also a matter of practical necessity. The integration of AI into digital public infrastructure (DPI) systems, such as reliable and secure biometrics in mobile devices can accelerate the accessibility